Ladies and gentlemen, please welcome President and Chief Financial Officer, Oracle, Safra Catz.
Good afternoon, everyone. It's wonderful to see you again this year. It's been an amazing Oracle OpenWorld. Unfortunately, most of it I've missed because I've been home sick in bed. I apologize for a couple of things in advance. First of all, my voice, the nonstop coughing I would do if I stood up here the whole afternoon, and the fact that we'll be switching a few things around today because of it. First, let me just thank you. Let me thank you all for being here, for spending the time with us. I hope you've gotten the chance to really spend some time at Oracle OpenWorld. This is, without a doubt, the most exciting OpenWorld we have ever had. We're going to share with you today really the fundamental of our future and our present, which is what makes it so much more exciting.
I always try to figure out where we are in our life cycle, I really do feel confident that I can actually say we are finally at the end of the very beginning. We've got just all the parts, and all the parts engineered to work together. I hope today we can really share with you our strategy for the future, our goals, our abilities, our products, and really our future results that will come from bringing it all together. Because my voice is just not going to be able to hold out, I'm going to come back and do Q&A after Mark's presentation, we're going to reorganize a little bit. Ken Bond is going to do the section I usually do, which includes our financial results, and he'll be followed by Mark, and then Mark and I will together take questions and answers.
With that, I'm going to pass over to Ken, and thank you very much.
Thank you, Safra. Thank you, Safra, also welcome all of you. Obviously, as Safra pointed out, it's been a very busy week here at Oracle OpenWorld. Just do a little housekeeping, then we'll go into the financial summary. If you haven't already grabbed a lunch, they're right outside. Bathroom is right here to my left, your right. The Oracle Wi-Fi is up and running. It's Oracle FAM, F-A-M in caps, so it's all there. You all should have electrical, we'll do that. This is my favorite slide that I always get to do. I have to read some language here, please indulge me. As a reminder, today's presentation will include forward-looking statements, including predictions, expectations, estimates, or other information that might be considered forward-looking.
While these forward-looking statements represent our current judgment on what the future holds, these statements are also subject to risks and uncertainties that may cause actual results to differ materially from statements being made today. Throughout today's discussion, we'll attempt to present some important factors relating to our business, which may potentially affect these forward-looking statements. As a result, we caution you against placing undue reliance on these forward-looking statements, which reflect our opinion only as of today. As a reminder, we're not obligating ourselves to revise or publicly release the results of any revision of these forward-looking statements in light of new information or future events. We encourage you to review our most recent reports on forms 10-K and 10-Q, and any applicable amendments for a complete discussion of these factors and other risks that may affect our future results or the market price of our stock.
Also would like to make sure everybody understands that the presentations today are for information purposes only. Lastly is, as a reminder, we will be using non-GAAP measures throughout the performance, especially the one I'll be doing shortly here. I just want to remind everybody that the GAAP and non-GAAP reconciliations are available on our website. Thank you for the indulgence. What you're going to see is a number of the speakers, as they're presenting today, will show you the safe harbor slide. They'll make a quick reference to it. When they're making that reference, they're referring you back to this portion of today's webcast. Okay, let's take a quick walk through the agenda. As you heard from Safra, we've made a couple little changes here, welcome. What we'll do then is we're juggling things a little bit.
I'll walk you through quickly the financial summary. Mark will talk with you then about the strategy. We'll bring Safra and Mark back up on stage for a Q&A. We'll start getting into the technology. We'll have Abhay Parasnis talk about the Oracle Cloud, the platform level. Steve Miranda will talk about Fusion and Software as a Service at the application level. We'll talk about our Engineered Systems. Juan Loaiza will speak to you about the Exadata products. Hasan Rizvi will speak to you about Exalogic. What we'll do is, in order to be a little bit more efficient about it, we'll bring Thomas up to address the Q&A for those products. We'll get into a break, that'll take us to about 3:00 P.M., just so everybody has an understanding about when that'll be. We expect to be there at 3:00 P.M.
We will get into the second half. Andy Mendelsohn will come up and talk about Oracle Database 12c. We will also continue with Engineered Systems. Balaji Yelamanchili will speak to you about Exalytics. Thomas will come up in data. We will do a Q&A on that portion of the day. Larry will be coming in to do a Q&A with you to wrap up. We expect the event will wrap up somewhere around 5:00 P.M., thereabouts. Okay. All right, let me take you through some of the financial summary. Safe harbor I just referred to. I want to hit four things today. First of which is a quick update on our financial performance over the last quarter and the last year.
I thought it would be good to touch on the business model a little bit. We will get a little bit deeper into some of our historical financial results and, of course, how we stack up. Most recent quarter reported was fiscal quarter Q1. We had 11% constant currency growth. These are constant currency numbers when I talk about growth rates. 11% constant currency growth, that compares to 11% the prior year. Double-digit net income growth and earnings per share growth on top of the same last year. Very strong cash flow, $5.7 billion this most recent quarter. Last fiscal year, which ended in May, 9% new software license growth. That was on top of FY 2011 that had 19%. Revenue was 4%. That is on top of 30% the prior year in FY 2011. We have seen margin expansion. We are back to record levels for last year.
We saw margin expansion in Q1. Double-digit earnings share growth in FY 2012. That was on top of 30% growth in FY 2011. Free cash flow continues to be very, very strong. In setting a baseline for you for a lot of the slides that come forward, Larry made a comment back in 2005 talking about our expectation that we could see earnings growth in excess of 20% over the following years. How have we done against that? Actually quite well. We have seen EPS growth averaging 20% over the last seven years. What is driving that? Big surprise to folks, is the software business does extremely well. Just putting it in perspective here, 87% of our margin dollars is being driven by software business, software new license, software support. Very strong part of our business. What makes that so strong? It is the fundamentals of the business model.
When we sell software licenses, new license sales, substantially all of those new sales come with a support contract. The renewal rates on those contracts, very high, very steady, very consistent. That leads to an installed base model on software support that, as you can see, has just gotten bigger and bigger when we look at our software sales overall. Started at 49%, most recently, 55% of our software revenue is from software support. That has been expanding. That business, by the way, comes with very healthy margins. Let us take a look through at some of our financials over the last few years. Top line growth, very solid, 18%. New software license, double digit. That growth over the last eight years, seven years, excuse me, has been very strong and it has been balanced. If we look across our geographies, all geographies showing double-digit growth.
Software support, big part of that story. It's an installed base that just keeps getting bigger. 17% growth over the last seven years. In excess of $16 billion in software support. These are big numbers. Now, we've been investing. Safra talked about there's probably never been a better time for Oracle in terms of our product portfolio. Never been stronger. Investments that we've been talking with you about over the last few quarters, making those investments in sales resources to start driving value from the wonderful products that we're bringing to market. The investments haven't been just in sales. We want you to understand, we're always continuing to innovate. Not only do we invest for growth, we invest for innovation. 4,000 people over the last eight quarters added to the R&D organization. How does that stack up with everything else?
94% of our headcount adds over the last two years have been sales. That's sales, not sales and marketing. It's sales in R&D. We're investing for growth. We're investing for innovation. Given the size and scale of Oracle, we have natural advantages. We're basically taking the development dollars, $4.5 billion last year, higher than that, approaching $5 billion this year. We're spreading that cost over an enormous base of 390,000-plus customers. That translates to margins that are best in class. I think we probably have the single best margins out there, especially for a company of our size. You can see here, this graph is just showing, depicting here, the smaller the company, the lower the margin, Oracle being large, better margin. You've seen that play out through the financials.
Margin expansion over the last five years, 500 basis points higher than they were back in FY 2005. Put broke out, you can see FY 2010, we broke that out in two halves for you. Second half, in the middle of the third quarter, actually, the last month of the third quarter, we acquired Sun Microsystems. We laid a hardware business on top of a software business, we're still 500 basis points higher. We're still expanding margins. Last year's 46% was back to record levels set in FY 2009. That starts to flow through as we work our way through the P&L. Operating income up 20%. In turn, drives a lot of cash. Our cash flow last quarter, $5.7 billion. We're putting that cash to work. We reinvest through R&D, the $4.5 billion last year, higher this year. We invest through M&A.
We also are in a position to generate sufficient cash to push back to shareholders $25 billion over the last eight years. Roughly 30% was just in the last fiscal year. If you look at the numbers, you can see that last year between buybacks and dividends, we were almost the same level as the prior three years combined. Very healthy cash flow generation. That flows to the bottom line again, 20% earnings growth. That's us. How do we compare? We took a look and we said, "Why don't we compare ourselves against some of our friends?" You've got some high flyers in here, fast-growing companies, and you have more steady mainstream companies that have been here for quite some time. Some having more recent challenges, some not. The point is, you've got some high flyers and you've got some of those others as well.
It's a balanced portfolio. We took a dozen companies here. We also looked at the Dow. To be clear, we took the Dow less the financials. The business model for financials and banks is different than others. You can see the names there. Anything that's financial service related, we took that out. Everything else, fair game, we threw them in. What do we do with it? We took our revenue and we normalized that revenue in FY 2005 to 100. We indexed it to 100. We aggregated all of the revenue for all of our tech peers, and we normalized that to 100. We took all those Dow 30s, excluding the financials, and we indexed all those back to 100. We laid it out. How have those numbers changed over time? We did pretty well on revenue.
We did pretty well on operating income. We did pretty well on net income. We do well on free cash flow. Guess what? We do just fine in earnings per share as well. Very strong, consistently outperforming the group of peers and the Dow 30s. A lot of stuff has happened. We took a look at how has our multiple been. We looked at the Dow 30. The world has changed a lot since 2005. We all remember 2008 and 2009. There's been compression, and we've all seen that. The Dow multiple has dropped 26 points. The tech peers, the companies we're looking at, they've gone down 33. Oracle, down 40. We're not trying to draw any conclusions, but these are the kinds of things that we look at sometimes when we do scratch our head a little bit.
With that, I'm going to turn it over to Mark. Mark?
Good. I'll be without a job. Hello? Hi. Safra's on injured reserve. She's going to be well enough. We'll do Q&A together. Hi. When I saw my schedule for the week, I have to tell you, this was the highlight of it, to come see all of you. I'm glad I've got her laughing. She's had a rough week. I'll tell you three things. First, I thought I'd tell you a little bit about strategically what we're trying to do. Second, tell you a little bit about our field organization, since that seems to be a hot button that everybody has. Tell you a little bit about the announcements we made at the show and how relevant they are, I think, to our customers. Safra and I will do Q&A, and I'll try to move through this at a rapid pace.
First, strategically, four things we're trying to get done. No change to our best-of-breed strategy. Everything we do in engineering, we do to win by the various point product. We want to be the best at what we do. We line up from engineering to sales to execute that strategy. Doesn't matter where at the stack, whether we're at storage, servers, OS, Database. You look at our progress in Linux, our progress in Database, the popularity of our Database, popularity of our middleware applications. We do have customers that want to buy from us in a heterogeneous environment, and we sell to them in a heterogeneous environment. We want to be open, high level of enterprise fit. No change to that strategy. What we do then, second, is we integrate these best-of-breed capabilities together.
One of the reasons, as we talk about Engineered Systems, that they're so attractive isn't just the integration, it's the fact that each individual piece part is in and of itself the best in class. That's why you get this huge exponent of performance and total cost of ownership as you do it. Engineered Systems manifest itself in Exadata, Exalogic, Exalytics, SPARC SuperCluster. They deliver extreme performance, and there are multiple benefits that come out of these Engineered Systems. One, they're just simpler. We provision the stuff, the customer doesn't. Second, they perform better. As Larry talked a lot Sunday night about not just performance, but performance and cost, if you will, TCO, is actually one conversation. We've historically talked about it being 100x faster, about being Exadata. The fact is you can translate that into cost savings.
In addition to TCO, performance, we can support them better. One of the things we announced June 6th, we announced the cloud, but probably got less publicity than the cloud, was platinum services. Our ability on certain configurations now to give 5-minute SLAs. Frankly, we can do better than that. But 5-minute SLAs when we see an SR. Highest levels of support you see in the industry. Third for us, you're going to hear more about that today, is this whole movement to Fusion Cloud Applications. It's a big transition. We're the only people, we're the only company with a suite of applications available in the cloud, but they're also available private cloud and available on-premise. Huge differentiator in the market. Brand new code. Salesforce circa 1997, 1998. SAP, nothing in the cloud of consequence. Highly differentiated position that we've got.
Fourth for us, we're taking things to market through industries, trying to solve industry problems, banking problems, healthcare problems, et cetera. That's what we're doing strategically. Now, last year, we added a lot of people to the field. We added them, and I have even better news, they're still here. Keeping you entertained over there, is it? We've had lower attrition than historically. We've hired. We've been able because of various reasons to attract, we think, very talented people. Our pipelines are up accordingly. I'm going to repeat again because every now and then I'll get a note that somebody says the Oracle field is in some transformation, some transition. I want to make sure I'm clear with you. They're not. They're not. Every salesperson at Oracle knows who their boss is right now. Every salesperson at Oracle has a territory.
Every salesperson at Oracle has a comp plan. They're being measured on their pipeline and their results accordingly. What we've done is while we've added more people, we've actually specialized them further. One more time, there's a big belief. Let me put it another way. I have a big belief that when you give a salesperson two things to sell, they self-specialize. We specialize for them, get them focused. We want them to be experts at what they sell, better than their competition. Simultaneously, we use a key account director process where there is a lead in the account amongst the group to align the total portfolio of Oracle to our customer. Large number of resources in the field. To Ken's point, I want to make sure I clear up one thing.
When we say we added in sales and R&D, there's also what did you add in sales? Because sales is a broad category. What we added in sales were two things, primarily quota-carrying sales reps and pre-sale technical reps that support the quota-carrying sales rep. In addition to that, we reduced the number of double-pay people in our quota-carrying sales organization. I don't want to get you into too much complexity, but just so you understand the transformation we've gone through. That's over. Focusing more people with a singular quota, further specialized in front of the customer who sell, supported by technical resource that's also expert, if you will, deeper technically in the product set aligned with those sales folks. When you put a label of sales, you can find operations people, alliance people, all types of people. That's not what we hired.
Quota-carrying and pre-sale technical people supporting those folks. At the show, some pretty big announcements. You're going to hear about them during the day. We announced a new Exadata. Is Juan here? Yeah. See, Juan hates this because I'm going to have to tell you about this thing so that when Juan talks to you about it's going to feel like, "Didn't the other guy tell me about this?" This is pretty big stuff. I used to be in the analytics business many moons ago when a three or four terabyte data warehouse was a big deal. Five terabyte, 10 terabyte. This thing has 26 terabytes of memory, four terabytes of DRAM, 22 terabytes of flash. Translated to, it's really fast. You've got certain jobs, and we talk about jobs with the Oracle Database that we can run 100 times faster with Exadata.
Certain jobs with this Dash 3, we could actually accelerate 20 times as fast as that. Fast. Big announcement for us. Remember, as I just went through that, I gave you that example with pure speed. You could do exactly what I said earlier, translate that into TCO for the customer. Second, we announced infrastructure and platform as a service to supplement our SaaS offerings out of the cloud. Third, we announced Database 12c, and the c being cloud. Again, he's going to talk to you about that. Lastly, we announced a thing that probably has got about as much buzz as anything I've seen for a while. We announced really the capability of the Oracle Cloud as a private cloud for our customers, meaning that you can now get that very same Oracle Cloud, and we'll put it behind your firewall.
We'll make it OpEx for you. We'll let you share resources with the Oracle Cloud while you do it, so you get elasticity as you do it, and forms the basis of what can be a fantastic consolidation engine, infrastructure as service. You get all the benefits of Exa, standard configurations, fully secure behind your firewall. We'll make it OpEx. We'll manage it for you. Four big announcements at the show that supplement what I tell you I think is probably, no doubt, the strongest portfolio in the industry, top to bottom. That's what we've done. With that, you want to come up, Safra? We'll just do Q&A?
Okay. By the way, those of you who want to ask questions for Larry, save your questions because he's coming late in the afternoon. He'll do basically just a Q&A. I tell you, these lights are so bright. If you're raising your hand, we can't even see it. Maybe we could turn them down. Adam?
Adam. Hi. Thank you. It's Adam from Morgan Stanley. The last couple of days, we've talked to a lot of partners and customers who are talking a lot more about Fusion and building pipe for Fusion, customers live on Fusion. At this point in the Fusion cycle, how do you size what you think the forward license revenue or opportunity set is, and how should we expect to see Fusion really start to impact the numbers over the next 4 to 8 quarters?
Listen. We've been working on Fusion for seven years, okay? This is the most modern product line, the broadest product line in the world, okay? We're pretty optimistic. Now, I'm not going to give you numbers of where this is going to go, but we're talking about over 100 modules. I pulled myself out of bed to meet with a couple of customers, a few customers, and they were talking about going live on Fusion Procurement, on all the different Fusion HCM. This has been really very, very warmly received. It's gone really well. We're very optimistic. I don't have a number for you to plug into the spreadsheet, but we think this has been a long time coming. We're very excited about it.
I mean, listen, we're putting a lot of energy behind it. I won't give you any numbers either, particularly if she won't. I think we're out aggressively selling. I think, Adam, you see it in the show of HCM now is great. We've got references. We've got live customers. We think we have a better solution than our competition, measured in just pure functionality. We've got more release coming. We've got salespeople that are now educated and trained. The run we're having in Europe and globally is exciting. We're pretty motivated about where we sit. Sales automation, as you know, was released after HCM. We're going through that exact same process with sales automation. We've got our organization aligned as of June 1. It's very focused on it. We've had some good wins. We talked about that at the quarter conference call.
We're pretty motivated by the position we're in. To Safra's point, these competitors we have are point product competitors. To the point she raised, we're the only one with a suite of capability, which I'm telling you, long run, as much as many of you want to talk about these little boutique players that are out there, companies not only have to look at these things vertically, they have to look at them horizontally and how they work together. Being the first mover with cloud SaaS apps on a suite of capability is going to be a huge differentiator for us out there. I agree with you. The reception, I think we're buoyed by the positive reception, and we're not thinking about it about growing this versus this. That's not how we think. We think about just growth. Just go get deals. Go win.
We'll take one.
Whatever. Somebody does, sure, I'm ready for anything.
Hi, it's Brent Thill from UBS. Safra, you achieved your 20% operating margin goal, 20% earnings growth goal. Can you just give us a sense of how you're thinking about that goal going forward? Is that your aspirations going forward? For Mark, just on Adam's question on Fusion.
We're doing two parts now? You get to do two? This will take forever.
I'll leave it there.
Okay.
Okay. Listen. We have gone into an area now where we are very focused on organic growth. We think we have all the best, most incredible parts. Those of you who may have watched Larry do some Q&A on television or in other venues here have to hear from us that we are extremely confident about the different product lines we have and their opportunity to really grow fast. We've invested massively in our distribution channel. That should tell you right there and then how optimistic we are. We are literally leaning in with the best product line, top to bottom, that we've ever had, and the reception has been incredible. In all the parts of the business that really matter, that we really look at very carefully, we believe that this is our chance to move forward.
We've got the parts, they're engineered to work together, no one has what we have. You can't stand up a single other company that's got the platform, the products, and the infrastructure that we have. People look and say, "We're new to the cloud. We're old world, and now we've come to the new world," Brent. That's not true, you see? The truth of the matter is Oracle always actually lives in the new world, okay? Larry has always led us with a vision. Maybe it wasn't called cloud seven years ago when we decided that we were going to build the most modern toolkit in the world to build the most modern, SaaS-ready applications in the world.
Maybe it wasn't called, I don't know, cloud when we decided to make sure our hardware and software can work together, and it can work either at our place or at our customer's place. Let me tell you, we've got everything for the cloud, and we think this is going to be an incredible next five-year run. Yeah, for the last seven years, it's been 20% compounded annual growth. There's nothing standing between us and doing that again with what we have now. Okay. You made it.
Thanks. Over here. Mark, Rick Sherlund. I'm just curious on the references for Fusion. Last year at this time, Larry announced Fusion. I think you had some difficulties probably selling it before you had good references. I guess the first part of the question is, where do you stand now in terms of real referenceable customers for that? I'm also curious about the Fusion behind the firewall on the private cloud. It seems that a lot of big companies are a little reluctant to put their apps out into the wild. Do you have any customers that are running it now in the Oracle private cloud? When will we begin to see that, and what kind of feedback are you getting from big customers in terms of their desire to have it behind the firewall versus, to say, Workday out in the wild?
Okay, Rick, thanks. Just one quick point of decorum, right? This thing is called Q&A as opposed to Qs and A, anyway. I'll try my best to get through the stuff you asked. One, you're right. Listen, it's much easier to sell. When you sell, you have to go through a process of provisioning the solution, getting them live, and then getting them referenceable. To your point, when we announce, it's obviously easier as you go through that process. In HCM, we're out there. We've got people that are live. We've got references. We've got references of extremely sophisticated logos, if I can use that term. We're out selling. We're in great shape. We're going through that same process now going live in sales automation. Right now in service, we've got more references than anybody on the planet. We've got the leading solution in service automation today.
I could go through things piece by piece, but we're in pretty darn good shape from a going live referenceability perspective as we go forward. To your point on the private cloud, you used this term "putting their apps into the wild," let me say it a different way than you said it. I think what we've done with this announcement is eliminate most every concern that an IT organization or a business would have regarding security, regarding anything outside their firewall, whether it be regulatory or just, "I'm just worried." We take that away when we deliver this in the private cloud. I want to separate two different things. Somebody running Fusion in their private cloud is different from what we announced Sunday, which is our delivery now of infrastructure capability. Infrastructure as a Service has the same infrastructure that which we run the Oracle Cloud.
We have customers now deploying Fusion. Remember, Fusion's got a couple big differentiators here. Fusion shows up by module. You can buy Fusion, you can buy a performance management app in HCM, you can use that in concert with your PeopleSoft application. You have the ability. You don't have to rip and replace this stuff. It's modular. It can run in the Oracle Cloud. It can run in a private cloud. It can run on-premise, you can change your mind. With what these product announcements, as Safra described, we've made it actually more complicated for the customer because we've taken every objection away. I mean that sort of tongue in cheek. Now it's to the fact that you actually have to think through what's best for you because we're going to do it for you whatever way you want us to do it.
John DiFucci, I see you. Please go ahead.
Thanks. It's John DiFucci from J.P. Morgan.
Sorry.
For the sake of now, you're Joe.
I'd go with Joe if I were you.
That's okay.
Sorry, John.
That's okay. Thanks, Tom.
I'm hilarious. No problem.
Mark, you mentioned that the sales team is measured on pipeline and performance. Performance, I take to mean actual sales. I'd expect that more recently anyway, it's more about pipeline because given the high hiring you've been doing recently, there's a lot of relatively newer people out in the sales force. When do you expect that to become more proportioned or balanced to historical norms, where you're really looking at actual sales versus pipeline? You've talked about this in the past, but I'd like to hear, as this progresses, when do you expect to start to see that in actual sales? You're obviously always looking at pipeline. You're bringing in new people. You're selling to the enterprise. It's six to nine months.
I really appreciate the way you continue to repackage the question as you go. Listen, I think you're asking when, basically, are these folks productive.
Exactly.
At the end of the day, we have people who have feathered into our model at different points, and they're at different points of maturity. We have people now that have been, if you did a sort of a starting date of June 1 of last year, you roll back 15 and a half months. Those people today are measured on sales. It's been enough training, assimilation time, it's sales. We have other people that have been here less than five or six months. They're being measured now on pipeline. It depends on tenure as it relates to what we described. Listen, John, we're focused on selling. We look at pipelines. We measure pipelines. At the end of the day, it's about deals. It's about transactions. It's about growing. That is, in the end, the core metric.
All you get with pipelines is, frankly, early indicators of interest. Frankly, pipelines growth in the absence of a conversion to a sale and understanding the different elements of conversion. As you know, those are different by product line. They're different by geography. In isolation, it's not a perfect science. You have to get into a very atomic level of detail to understand how all this is going. That said, as I said on the last conference call, our pipeline is up materially as a result of what Safra described a few minutes ago.
I guess, just quick follow-up there. The bulk of hiring was done, I don't know, six months ago, something like that. Should we expect that pipeline to transfer into sales?
Let me go one more time. We expect to have a good year. I'm not going to give you I know it'd be great to have a call, and we can take you through each guy and tell you how they're doing, et cetera, et cetera, and how their pipeline's growing, and I'd be glad to do it. She's told me I can't do that because I really told her, I said, "That'd be great. I could just take you through region by region, person by person, how their pipeline's going." She said, "Don't." Listen, we feel good about who we've hired. We think we now have hired a lot of people. We think we've hired good people. We've had them in front of a customer. Our pipelines are growing. They feather in over multiple quarters.
We did do some hiring of Q1 of last year and Q2 of last year and Q3 of last year. We've done hiring in Q1 of this year. We've made a major move in our business analytics business. That's a business actually we don't talk about much. We feel great about our position in business analytics. Unfortunately, we feel so good about our other positions, we don't talk about it as much. I just did a keynote, I don't know, three or four hours ago. I don't know if Balaji's in here. Are you in the back? Balaji and I did a keynote on just the subject of business analytics, and we're hiring business analytics right now. Will those people come online? Some may be a little bit pipeline in Q2, hopefully business in Q3.
Really, if you look at our industry businesses and our regions, we expect a good year. Anybody? Hey.
Brad Reback from Stifel. I'll make it very quick.
Good.
Safra, as we think about the move to the cloud and the CapEx requirements for the business, do they increase meaningfully?
Interestingly enough, they're not nearly what you would think they are. For us, when looking both at CapEx and at margins, generally, it's all very manageable within our margin envelope at this point. Luckily, we both make the software, make the hardware, so I get a reasonably good price for myself. Then additionally, because we do it in scale, because we also write the software, we reduce the one thing that usually actually costs a lot more than the CapEx and even more than the energy, and that is the labor. That's really what must be bogging down many of the other smaller cloud vendors, between the labor to manage these systems and their sales and marketing expenses, which they just don't have the kind of scale we do. We don't see CapEx as an issue or even energy costs or labor or sales and marketing.
We actually think we're going to be able to run the cloud at the kind of margins that I think will make all of you reevaluate what all the little guys are doing, or even the midsize guys. We think it can be run very well. I know I've answered more than just your question, it's all within our margin envelope at this point.
We also bought this company called Xsigo. Every time I say it, I'm told I say it wrong. Xsigo. I'm sure someone will tell me it's wrong.
Xsigo, yeah.
What we do in that product is we actually provision more efficiently network connections. Add that to everything Safra said, we lower our network provisioning costs dramatically. As a result, I'll just say it one last time, what a huge advantage when the major supplier, almost the only supplier of everything we do to run our cloud operations, is ourself.
Oh, okay.
Joel, you had your hand up.
Thanks, Mark. Joel Fishbein from Lazard Capital Markets. To follow up on the new Exadata infrastructure as a service and Database 12c, can you give us when will they be GA, and where are they in that cycle? That would be helpful to get clarity on that.
Exadata.
X3.
Yes. They're out, man. I'm making them right now. Orders, they're shipping. They're going. I'm out of X2s. All I got is X3s.
The infrastructure as a service pipeline and Database 12c.
That is actually, one of the technical guys can tell me exactly when Oh, Andy, the cloud infrastructure as a service is?
12c. 12c I see you.
Oh, 12c, we're not telling you yet because it ain't ready. We will tell you the minute it's out. The minute it's out. It's coming out soon.
Infrastructure as a service this calendar year.
Yep.
Okay.
Hi, this is Justin Bandy from Artisan Partners, over here. Safra, I'd like to hear your thoughts on capital allocation going forward. In your comments, you said you're happy with the pieces that you have and your focus is on organic growth. Does this mean we'll see a different pace of M&A going forward? Your thoughts on dividend and buyback, what should we think about that? Thanks.
Sure. Well, it's like the past is prelude. What we're going to always do is, first of all, our first rule is we're not a bank, okay? It's your money, unless we've got somewhere great to use it, we're sending it home. Historically, that's what we've done. As you see, our buybacks have been really large lately. In regarding a dividend, our board constantly is looking at that. Obviously, there are potentially tax law changes coming up, all of us are looking for guidance. We're going to have to see what plays out in the next few months as to how that goes. We're constantly looking at it, the board makes that call. On acquisitions, you know us.
I'm a personal shopper for our CEO, when we find something that's really compelling that we think we can make a lot of money for all of you with, we're going to buy it. We don't feel pressed to buy anything. We've got all the most incredible parts right now. You can just hear it in sort of every part of Oracle OpenWorld. Every once in a while, there are things that come up that are a perfect fit, to be a perfect fit, they've got to match what we're doing. They've got to pull through other products, they've got to fit within our strategy. We will look at things. We still continue to do acquisitions, I'm not telegraphing one or the other. In my comment earlier about we've got the greatest parts, we do.
If something else comes up that we see that's compelling, we're going to look at it. For us, it's all about the return, it's all about growth, and it's ultimately about sending the capital back to the people who own it, which is you guys.
Hi.
Yeah.
Hi. I'm short. People can't see me.
It's so bright. You guys can't even realize that we can't even see you.
I'm sure. Laura Lederman with William Blair. Thank you.
Yes.
As you acquire products, like RightNow, where those vendors had them only running at their site, that shifted from running things on-premise. When you acquire companies, will you allow the customers to run it on-premise? When you buy pure cloud companies, the same type of choice you give them for the products you create yourself? Is that going to be a policy for everything you acquire-
Well-
even if it's pure cloud, that you will allow them to run it on-premise?
It's going to depend if that's a business model that makes sense. Depending on what the service is, and whether it makes sense and whether it's technically feasible and manageable. I think we'll have to take it on a case-by-case basis. As a general matter, we do believe in customer choice. We want to give them the opportunity, and we want to make sure, though, that to the extent that we give the opportunity, we're only going to do it if we think there's a market for it.
Yeah. Listen, what you described is our strategy, but you're going to have to take each of those products that have been architected for the cloud. We have typically, to your point, bought SaaS products, point products to fill in some parts of our portfolio, and those products work in the cloud. We'll always evaluate whether it makes sense to ship them, just as we would make decisions on how we integrate products across our product line when we bring them into the company. Those generally are SaaS products today. Nothing would stop us from running those as a private cloud solution from a technical perspective. Just a time check.
Yeah, we don't.
How, Tim?
Five.
Okay.
Okay.
Thanks.
Hi. Michael Turits from Raymond James.
Hi.
In sales specialization, can you talk about in what segments you're specializing cloud sales from non-cloud sales? How do you incent people to sell cloud when conceivably those could be smaller upfront deals?
Yeah. We actually don't think of it the way you described. We don't have a cloud sales force and a non-cloud sales force. We organize our sales force really by function, by process, by buyer. We actually have an HCM sales force. And in the HCM kit, you have on-premise, you have private cloud capability, you have SaaS capabilities. Our sales force is very incented to sell SaaS cloud. They're incented against what's determined ARR, annual recurring revenue, and then they get a factor against that ARR, and it's very attractive compensation. I think you'll find our compensation plan. You may want to apply for a job. It's a very lucrative assignment. That's provided you sell. Yes. That's great. Send it to me. Email would be great.
We do incent them by ARR, and they are highly incented to sell in the cloud. Yes, I see you.
TV Ramamurti, Pazina. Infrastructure as a service. How are you thinking about data as a business? Why is that strategic for you? You can see that why SaaS is very important for you, why platform as a service is also important for you. What are your aspirations from infrastructure as a service?
Boy, I want to make sure you get this one, so you get context for what our customers are dealing with. Our customers have incredible data growth. If you talk to our customers, and I hope some of you did here at the show, our big customers are averaging 35%, 40%, 45% data growth. Okay? We have customers, if you go to banks, any of the big banks in this country, you're going to find they have 8,000, 100 petabytes of storage. They're paying $8,000 to $10,000 a terabyte to buy storage. Let me just connect the dots on that math for you. If I had 100 petabytes and I grow by 40%, I'm going to buy 40 petabytes. If I have to spend $10,000 a terabyte, can somebody do that math? Safra can . Hundreds of millions of dollars. Huge. This is a non-sustainable model.
When we apply Infrastructure as a Service. By the way, many of those customers, the most popular database they have in their business is Oracle. A very large percentage of that data sits in Oracle. That data, that opportunity for us to move that to the infrastructure we're describing, to Exadata, just as an example, gives us an opportunity not just to run it faster, but one of the key things when Juan gets up, and I hate to give more of Juan's material, he always loves when I present for him. We've got compression capability. We now can compress that data in some cases 10x or a 10th, meaning, let's just say, hypothetically, I had 50% of my data in Oracle. You can insert 60%, 70%, 80%, whatever number you want for your algorithm. Meaning, I had 50 petabytes in Oracle.
If I could compress it 10 to one, the 50 turns into five. Now my remaining 50 plus my five turn into 55 from 100. Now I grow 40% from that. I just saved hundreds of millions of dollars. I just gave you one example of a consolidation thing that's very real around customers that have Oracle. It is a huge opportunity for us. Huge. We've now taken away all of the concerns related to doing that via the cloud by being able to do that as a private cloud Infrastructure as a Service.
Okay, I think that's it for us.
All right.
And-
Thank you.
Thank you very much.
Thank you.
Ladies and gentlemen, please welcome Senior Vice President, Product Development, Oracle, Abhay Parasnis.
Yeah, you were great. Yeah.
I'm going to spend the next 20 minutes or so talking about Oracle Cloud. My goal is to give you a sense of the overall strategy, as well as what we believe are some of the unique and core differentiators to our cloud strategy, all the way from infrastructure as a service, platform as a service, and software as a service. I'm going to do this jointly with Steve Miranda, who's going to come right after me, and he's going to focus on the application parts of our SaaS strategy for the cloud. Okay. I'm just going to refer back to Ken, covered these two slides at the beginning, safe harbor, as well as the product roadmap direction. At a high level, what is our cloud strategy and vision?
It really revolves around three core pillars, actually, you're going to find a lot of what I cover is already covered by Mark. He thought he was covering Juan's material, but he's actually covering a lot of my slides as well. We have three core pillars, and they're very simple, but we believe they are uniquely differentiated in the marketplace. First, part of our cloud strategy starts with having the industry's most complete and comprehensive SaaS suite. Take our portfolio of applications and really deliver the entire suite of applications that we deliver today on-premises, deliver that in the cloud. The two core elements of our SaaS strategy are take the best-of-breed functional workloads and deliver them in the cloud, but deliver them in a fashion where they are completely integrated.
The integrated SaaS suite, we believe is the first element of our core strategy that's uniquely differentiated. Now, building on top of that, the second part of our strategy is to take the SaaS suite and deliver adjunct with that, a platform as a service suite that is completely integrated with our SaaS suite, that allows both enterprise customers, but also ISVs and partners, to take our SaaS suite and extend that with custom applications or other packaged applications around it. The combination of SaaS and the PaaS suite together, we believe going hand-in-hand when delivered as a complete vertical platform, is a very uniquely differentiated capability.
Just to step back, if you look at the market today, there are vendors in the cloud that deliver either infrastructure as a service or maybe platform as a service, but don't have as much of a SaaS application suite. Then there are vendors who deliver application capabilities, but either don't have an open platform or don't have any sort of infrastructure as a service capability at all. The notion of having this fully integrated suite is one of our core tenets for the cloud going forward. Last but not the least, is some of the announcements we made earlier in this week at Oracle OpenWorld, is we are going to deliver the full PaaS and SaaS suite on an enterprise-grade infrastructure in conjunction with our Engineered Systems in a fashion that allows customers the deployment choice.
Not only do they get the best-of-breed applications and the best-of-breed platform, but they get to choose whether they want to deploy those capabilities either in Oracle's public cloud, the one that we run and manage, a private cloud that's behind their firewall that is still owned and operated by Oracle, but it is running in their environment. Or they can take the exact same capabilities of our platforms and applications and buy them as licensed software to run on their own hardware in a traditional model. As we look ahead, I'm going to walk you through now the core capabilities, the roadmap, but more importantly, what we believe are the unique capabilities of each of our tiers of the cloud. To begin with, there are four layers.
There's a common infrastructure layer or infrastructure as a service layer that really enables us to give this deployment flexibility of going from public cloud, private cloud, and hybrid clouds, where the workloads actually span both the environments. On top of that, we have a fully integrated suite of applications that are delivered in conjunction with our platform technologies, the most popular database on the planet, and the most popular and widely deployed middleware on the planet. Delivering those application and platform services. Last but not the least is set of new capabilities we have added recently to Oracle Cloud in the enterprise social space. These are capabilities that extend further our SaaS suite into mobility, social, and big data. I'm going to walk you through now quickly through each of these layers.
The common infrastructure services layer is really a layer that we have been working on for a long time. It's at the foundation of our own cloud. This is what we are running our own cloud on in the Oracle public cloud. This is also the secret sauce, if you will, that allows us to take our solutions that we are building and deliver them both as public cloud, but also private cloud when combined with our Engineered Systems. This is the layer that one of the questions earlier was, how do we think about new applications that get delivered either in the public cloud, but over time need to move down?
It's this layer of abstraction that gives us the choice for customers to take Engineered Systems, Exadata, Exalogic, Exalytics, and these services on top of which then you can run all of your capabilities on top, whether it be PaaS or SaaS. To begin with, we have the core storage and virtualized compute infrastructure. We have an enterprise-grade secure identity infrastructure, and this is very critical. The identity infrastructure is one of those most critical elements to span the public-private cloud topologies. Then on top of that, we have set up capabilities for developers, things like messaging, queuing, notification gateways, network topology that Mark talked about earlier, caching. Our goal with this layer is to have industry's best from a performance as well as cost performance standpoint, infrastructure layer to run our global cloud, but also give the same capability to private cloud customers.
On top of that is our Platform as a Service layer. This is really a layer of capabilities that mirrors our strength and assets as a company in the traditional technology part of our business, which is database and industry's first kind of best applied middleware layer. At this capability, we have three core differentiators, we believe, as we take our PaaS into the marketplace. First of all, taking the most widely deployed and used database and Java middleware product, we are delivering our cloud in a fashion where all of the existing apps, all of the existing apps, which are hundreds of thousands of applications written by customers or ISVs, can be deployed without a single code change. They don't have to rewrite their applications. They don't have to reconfigure their applications. They can take those existing applications and move them to our public cloud.
We believe that is a huge differentiation and it significantly lowers the barrier for these customers to adopt our platform cloud. Second, we are continuing to invest and deliver a portfolio of platform capabilities that mirrors our platform portfolio on-premises. Capabilities like collaboration services, capabilities like BI and analytics, capabilities like big data as a service. We will continue to offer those so that the same partner ecosystem, same customers who use these today on-premises, have a seamless path to take some of these workloads into public cloud without rewriting their applications. Second part of our PaaS strategy that is unique is the entire PaaS platform is built on open standards. One of the biggest issues marketplace today has with PaaS platforms from some of the SaaS vendors, players like Salesforce, is their platform is a proprietary platform.
If you as an ISV or if you as a customer write your application to their Platform as a Service, you have no portability whatsoever of taking that application anywhere else. To contrast that, our Platform as a Service is written entirely with standards like Java, web services, HTML, SQL. Not only can you take your application and today decide to run it on our public cloud, six months later, you may decide you want to move it to your private cloud, or if you decide you want to just take it to some completely different vendor in the marketplace, you can do that. We don't believe a lot of the customers will do it, but there is no lock-in. Architecturally, the platform is architected to be open.
I want to give you a sense a little bit, switching from products and capabilities into what are we seeing in the marketplace today. What are some of the customers who are using our Platform-as-a-Service cloud? What are the kinds of applications and adoption scenarios we are seeing? Broadly speaking, we see four broad categories of early adoption momentum, if you will. At the bottom of the slide, you will see there are two workloads, which is the first category we see is large number of our existing enterprise customers have applications that they have written in Java. On top of our Database, that they simply want to drive a consolidation of the workloads, much the same way they have driven consolidation on virtualization, into now the cloud.
The first thing we see with a lot of our customers like Windstream, is they're taking their existing WebLogic-based Java application or existing Database applications written in Oracle APEX SQL and simply moving them as a consolidation exercise to our cloud. Similarly, we see a lot of our customers and more importantly, partners. This is another unique differentiation for us in our cloud, which is partners like Siemens and Accenture. They are seeing Oracle PaaS cloud as a very attractive model for doing things like using the cloud to do testing, development during projects, and then when the customers want to go live in production, move those workloads from public cloud to a private cloud for that customer behind the firewall.
This notion of scoped workloads, where they start in one place on public cloud, and as they go through the life cycle, they move it to private cloud or custom deployments. If the bottom two scenarios are mostly around existing application, the top part of the slide shows what we are seeing as new application development. Our PaaS cloud is appealing for both the scenarios. First is SaaS apps. There are a large number of new vendors and new ISVs that are looking to deliver new kinds of SaaS applications that want a robust enterprise-grade platform. The examples there is somebody like Emerson, who wants to get into now business of delivering a new SaaS service for doing data center energy management.
They looked at how they could deliver that service globally, and they realized that the cost structure, both OpEx and CapEx, of offering it on top of Oracle PaaS, is far more economical for them than trying to build that entire infrastructure themselves. Similarly, the last piece, which is customers who are taking our extensions on SaaS applications. Partners like CRMIT, they are taking our Oracle Fusion CRM, SaaS applications, and they are extending those with extensions that they are going to market with. This is a partner opportunity that is building on top of that virtuous cycle that I talked about, that SaaS extending to PaaS and more applications on PaaS means our SaaS applications get stickier and more useful. Speaking of SaaS, at the core of our cloud strategy is really having the industry's most comprehensive SaaS suite.
The goal for us is very simple, to literally cover every single mission-critical business process within the enterprise via an integrated SaaS workload that is both best-of-breed in its functionality in depth, but it is pre-integrated with all the other business processes that the enterprises have to deliver on. Steve is going to talk a lot more about this part of the presentation, so I am going to skip through and have him cover this in a few minutes. Beyond PaaS and SaaS, the other key part of our cloud strategy, the one that we have made significant strategic investments over the last six to nine months and have made substantial progress on, is a new layer of capabilities called social relationship management.
This is a layer of software services that we deliver that really integrates seamlessly with our SaaS suite, but brings newer experiences around social, mobility, data, and insights into your social applications, as well as your core LOB processes. The key approach we are taking with social, unlike some of the competitive landscape today, is today, a lot of the competition treats social as an either afterthought or a point solution. Most of the social applications, whether it be marketing, whether it be insights, are disjoint from the core LOB application processes. The thing we are focusing on is to deliver industry's first, and what we believe to be the only integrated social relationship platform. Where social is no longer a point solution, but rather it's blended completely and seamlessly with all of our SaaS applications.
For example, our social marketing capabilities are blended and integrated with our Fusion CRM sales and marketing. Our social engagement and monitoring products and services are seamlessly integrated with our customer service offerings from RightNow into Oracle Cloud. Social network, employee collaboration services are similarly integrated and blended with sales and marketing, talent management, HR, all of the business processes that customers are using today now just light up with social capabilities inside them rather than a separate island. In the interest of time, I am going to go through this quickly, but since this is a new area, I did want to touch upon, we are seeing already significant adoption and momentum in the marketplace with some of the largest brands and largest customers already embracing social in a big way, along with using SaaS applications from Oracle. Starting with marketing.
This is an example where somebody like Intel is essentially using Oracle's marketing offerings for social in a blended fashion to do multi-channel marketing, where they are taking their traditional CRM marketing and extending now social as one new channel. Similarly for monitoring engagement. This is a case where companies like McDonald's or Pepsi are using social marketing and social community management from Oracle to really manage their brand presence on a variety of social channels, Facebook, YouTube, Google +, et cetera. This is a case where we are seeing now a very interesting virtuous cycle build up, where customers who are using our SaaS applications are getting more value by simply turning on the social capabilities. As we go forward, you will continue to see us blend social more and more throughout our application portfolio rather than treating it as a point solution.
Of course, all of these capabilities are built. Mark talked about quite a bit of this at the beginning, I'm not going to touch a lot of the same things again. It's important to note that not only do we have the most comprehensive stack, all the way from infrastructure, platform, and SaaS, the way we have architected the entire Oracle Cloud footprint is exactly how some of our largest enterprise customers would architect it if they were building a private cloud. What that means is both at the data and the middle tier, we have created the proper isolation and sandboxing that no tenant or customer's data is commingled with another tenant.
The beauty of this model is this is one of those key things that allows us to go and offer now a infrastructure as a service and a private cloud offering, where a customer can start in public cloud. That's exactly how we have architected our public cloud. At some point, if they decide they want to move that workload from public cloud to behind-the-firewall private cloud, turns out, architecturally, both those things are built exactly the same way, using our Engineered Systems, using our database products, using our middleware products, and they are managed using the exact same set of Oracle Enterprise Manager capabilities that they would use if they were managing it on their own. This notion of using a completely symmetric hardware and software stack is extremely powerful for our Oracle Cloud going forward.
On top of that, we have a very simple, easy-to-use subscription model that makes the entire consumption experience of Oracle Cloud extremely simple, self-service, and a pricing model that is highly tailored to doing both per-user pricing for applications and a consumption-oriented pricing model for PaaS and infrastructure services. As I wrap up, one of the things we are excited about, both at this Oracle OpenWorld as well as as we look kind of the last 12 months, we are now starting to see a very strong and broad adoption cycle across some of the best-named, well-known customers across the globe. Not only are they adopting portions of our cloud, what's more interesting is they are starting to see the value and realize the value of adopting all tiers of our cloud, PaaS, SaaS, and IaaS.
In summary, here are the three things I would like to leave you with as we think about kind of the Oracle Cloud all up. First, Oracle is aggressively moving and delivering on a complete portfolio that we think is unmatched. Unmatched both in the full stack, infrastructure, platform, apps, and unmatched in the deployment choice of public, private, and hybrid cloud. Second, Mark talked about we are not delivering only a complete cloud, but each of our components are best of breed. The depth of the functionality in Oracle Cloud and the breadth via the integration is uniquely positioned both for our customers, but more importantly for the partner ecosystem that is also rapidly now converging around our cloud and starting to take it to the market.
With that, I'm going to invite Steve Miranda on stage, and he's going to walk you through the SaaS portion of our cloud strategy in more detail.
Okay, there's safe harbor again. Thank you, Abhay. Let me first go through the SaaS strategy. As Mark talked about earlier, he talked about removing the objections for the customers. I think of it slightly differently. Customers have particular problems, and those problems are bounded by a couple of variables. Those variables have time elements to it, they have sometimes technical elements to it, they have kind of functional choice elements to it. What we try to do is deliver, first of all, the most complete set of applications possible. This gives our customers not only a solution today in some cases, but in a solution going forward. Because while in some cases they want to solve a particular niche problem today, they want to know they have a roadmap going forward.
In cases where we compete against the best of breed, we develop the richest, deepest applications across a number of different product families and supplement those with choice. We integrate those product families together, the best-in-class product families integrated together, build out integrations not only with the organically built applications that we have, but also with the acquired applications that we've purchased. We exploit the platform as a service capabilities, and especially social, being a component into several different applications. In fact, I'll give you three examples where I think social is a fundamental aspect to applications these days, where you can't do without it. It has to be incorporated into apps. I believe it's going to be incorporated completely into applications, but fully essential in three different examples, at least today going forward.
What I'll try to illustrate is how we do this, give you some background on the SaaS applications, what we've built organically, how we've supplemented that through acquisitions, illustrate a number of customer examples. There were questions earlier about references. I picked the particular customer examples to showcase a particular customer problem, and therefore how they've exercised the choice and why either depth of functionality, integration, or breadth of suite has been critical to them, to solve their business problem. First off, what were the driving tenants? These have always been the driving tenants from day one of the applications that we presented in Fusion Applications that we started to build day one. First off, we always wanted to build a complete suite of applications.
Secondly, we felt that it's critically important to build across the boards on industry best and industry standard technology for ease of integration, for ease of extensibility, particularly in the cloud, for ease of interoperability going forward, and so that our customers, as they extend the application and have to live in an ecosystem of other applications, they have the best in class and lowest cost, lowest risk platform going forward, as opposed to proprietary platforms which tend to age, which means cost and risk. We wanted to extend that to the most modern user interface, including HTML5, which has given us the nimbleness to, though we started Fusion applications long before tablets, at least in the form they exist today, were available, long before the phones of today were available.
We've now quickly made all the Fusion applications available, and they all run on the tablet devices, both Android devices as well as iPad devices, because we went with a standard-based approach as opposed to a native or proprietary approach on a particular platform. We always built in, from the beginning, business intelligence. We thought that was going to be the key, and now as big data takes hold, that's even more of a key. We designed at the start what we initially called collaboration built in, which means particular event points within the application. Now, collaboration has evolved to be social, which meant that we didn't have to re-architect the applications now that social is much more pervasive, but we had the collaboration event points in the application built in from day one.
Now we can easily plug into Twitter and Facebook and Google+ and LinkedIn and other social platforms as they evolve. We built always for SaaS, but always with data isolation, as Abhay alluded to before. When we talk about the product families and the product modules, what do we really talk about? I'm going to go through these in a fair amount of detail as we go forward. We're talking about HCM, talent management. HCM, obviously, core records management, benefits, payroll. Talent management, things like performance reviews, compensation management, succession planning. Sales and marketing, things like entering leads, forecasts, quota management, territory management. Customer services and support. This is basically ticket logging, response, knowledge management to be able to respond to customers, and then e-commerce to be able to take it from a ticket into an upsell.
Financials, which is things like general ledger, accounts payable, accounts receivable, expense reporting. Procurement and sourcing and inventory, which is the supply chain components to be able to buy, manage suppliers, procure, and have the online sourcing. Project and portfolio management to track projects, to bill contractors, to be able to charge back for professional services type of fees. Governance, risk, and compliance for things like segregation of duties, Sarbanes-Oxley compliance, and a host of other similar regulatory requirements around the world. All built on the common infrastructure of Exadata, Exalogic, Exalytics, all built on a common set of platform services that we not only use but are available now as a standalone service to our customers, so the database as a service, the Java as a service, and all take into account the social components that we've added to the suite.
What it looks underneath, while there's a lot of attention paid with Oracle moving to the SaaS through the acquisitions, really what we have is a vast array of organically built Fusion Applications from HCM, sales and marketing, financials, procurement, project and portfolio management, and GRC, which is when they talk about the seven years of building the applications, that's that breadth of investment across those different product families and product modules that were all organically built. To make the journey to the cloud, we're not new to the game. We've been working on these applications for quite some time. Then we supplemented those applications with Taleo for talent management, specifically around recruiting and learning. Then we further supplement it with service and e-commerce with the acquisitions of RightNow, ATG, Endeca, and InQuira.
Basically, those components fit very nicely with extremely little overlap into the existing built application to give us a comprehensive, complete suite of applications. Again, the social acquisitions of Collective Intellect, Vitrue, and SelectMinds, I said earlier, it's a social platform, but it's a platform that gets embedded into the applications. It's not a separate standalone social app. Let me give you the three examples that I said earlier that I would argue are basically table stakes for the apps today. First, with SelectMinds for recruiting. Recruiting is basically if you have a job site, either on your internal website or you want to post to job boards, you want to do recruiting.
What social recruiting means is that if someone within your organization wants to take that job post and share it, either on Twitter or Facebook or Google+, and you want that to further be passed along in a social network. Basically use your own employees to do recruiting in their social sphere and then track that back. SelectMinds provides capability integrated with the best in class, the world's leading recruiting solution, Taleo, so that we now have the capability to take a job post, have employees post it to their social networks, have it tracked back so that employee gets referral credit, and you get business intelligence. A fundamental table stakes part of recruiting today. If you're doing recruiting, and you're not using social networks, you're probably missing it. Second example is in the RightNow case for service.
Service is what's typically what was called call centers, where people would have a question about your product and phone in. Call centers started to have multi-channels with just chat and email because people weren't doing phone calls. Today, what's happening is more and more is people aren't going to the brand for service. Meaning if you have an issue, you don't necessarily call or email the service provider. You sometimes post the question, or even worse, post a comment, usually negative about the service, onto a social network of Twitter, Facebook, Google+, et cetera, for either help responses or frankly just to vent that this is an issue.
Modern call centers or modern service agencies are using products like Collective Intellect, again, combined with the world's leading SaaS service provider RightN ow, so that they monitor the social networks to be able to respond to questions, inquiries, complaints that don't actually come into your call center directly, but instead go to the social network, and to be able to respond in the context that it's known. Second example of what I would argue is where social is now part of a table stakes application component. A third example is, of course, marketing. In Thomas Kurian's keynote on Tuesday, we demonstrated using social for marketing. Traditional marketing campaigns used to be you send out flyers, or you have programs or events. That migrated to email.
Now it's table stakes to have social as a component to a marketing event, where not only do you send your emails off, but you post advertisements on Facebook or Twitter or Google+ as a social component. It gives you better lead management because you have better data as far as who's coming in, and integrate that to sales. Social is key, not a standalone application into itself, though there's certainly uses there, but three examples where it's very rapidly become table stakes into the applications. A little bit more into what's in each product family, and then I'll get into some examples.
To be clear on HR, global HR, which is basically records management, compensation, stability to give raises, promotions, bonuses, incentives, benefits processing, and this is global benefits processing, so not only benefits through the U.S. but also the global with different regulations. Payrolls, that's a global payroll as well as U.S. payroll. Absence management. Network at work, which is another example of social integrated in. This is basically the ability for your own employees to have a Facebook-like page on your internal directory. Worker portrait, and then time and labor. For some customer examples. The first one is Elizabeth Arden, and the reason why I call Elizabeth Arden, this is an existing Oracle customer on HCM PeopleSoft, but they were basically a North America PeopleSoft customer with a set of disparate systems, meaning spreadsheets, niche systems, no real central HR system globally.
Here's where the depth and the migration was important to them. They wanted a quick solution. They are live today in core HR as well as talent management SaaS in Europe. They're going to go live in Asia Pacific later on this year or early next calendar year. Then finally upgrade their North America HCM system from PeopleSoft to Fusion HCM going forward. Kind of an example of what we call a coexist, running side by side to PeopleSoft, but important to them to have a migration path going forward. The second example I'll call out is UBS. UBS is in the process of implementing the HCM system. I call out UBS for a couple different reasons. They've selected us in SaaS for core HCM and talent management.
UBS, because they are a highly regulated company and because they have very strict standards as far as IT, chose SaaS. Before they selected us, they probably put us through the most scrutiny in terms of everything from the way we do data encryption, both across the wire and internally, the way we do data backup, the way the application is secured internally, our business processes for securing the application, our data recovery type of systems or disaster recovery type of systems, leveraging things like Database Vault in the Oracle Database. A whole litany of security requirements. In addition to that, the fact that they're a European bank and the fact that this is HR data with personal identifiable information, they did not want the data housed in the U.S. We've housed them in our European data center. Another example of our breadth globally.
Satisfied not only all their security requirements technically, but also with physical location, legal things, and so that was that example. The third example I'll call out is Brocade. This is a separate case of coexistence. Brocade is a big E-Business Suite customer. They run on-premise today, and they are continuing to run on-premise. However, they needed a rapidly deployed talent management solution. They deployed Fusion Talent Management in the cloud, both compensation and goals. Basically, they set their goals earlier this calendar year. They assessed their employees against those goals, and based on hitting those goals, they ran their compensation round. That data was fed from their E-Business Suite system with integration that we pre-built on-premise. In other words, the employees were fed into Fusion Cloud. The comp was issued, and then it's fed back into E-Business Suite on-premise for payroll.
As they move forward, they're looking to move more and more applications to SaaS eventually, but they're not ready to move all of the applications to the cloud today. They were able to get a quick need fulfilled but have a complete roadmap because of the breadth of suite we offer in the SaaS environment. Moving on to sales and marketing. A little bit more about, first, what's included. First is install base. What does the customers own? Territory management, assigning the territories and just kind of carving up the customer base to the different sales rep from a management perspective, so you can better optimize coverage of your sales. Multi-channel marketing, meaning be able to market, as I alluded to earlier, not only in email or call downs or traditional marketing campaigns, but also social. Partner relationship management.
This is for companies that go through resellers or sell through partners, to be able to manage them and give them access to the sales tools and utilities. Something we call Sales Predictor, which is where we leverage our BI technology. Sales Predictor basically mines the install base, and based on your results of actual install sales, predicts where you might have best opportunity to sell to new customers next. Which gives the sales reps more qualified leads because you're predicting sales based on existing references and existing data. Leads and opportunities, forecasting and quota management. In customer service, this is from RightNow, we have web self-service, so basically the ability for the customer to come online.
With the knowledge management capabilities that we provided at Oracle in addition to RightNow, have the ability for the customer to more rapidly find the issue that they have, to better answer their questions on their own. Meaning if you go to a company website to search for an issue on the product, how quickly can you find someone who had the same problem, the same resolution? This includes mobile of all types that you would expect, including tablet and phone, chat and browse capabilities, the ability to go onto the vendor website and chat via the browser. Of course, email, knowledge management I covered, to traditional contact center, and then support communities, which is another flavor of social built-in to be able to go and have message boards or forums, if you will, so that the users can interact with one another.
A couple examples here in sales and marketing. First of all, Key Energy. This was an example of a pure best-of-breed sales situation for sales and marketing. Key Energy chose Fusion CRM for sales and marketing. A couple key components. One was of Sales Predictor. The second was what we call a personal sales campaign, which is the ability for a sales rep to be able to start a marketing campaign specifically to their targeted customers, not have to have a global marketing campaign. The third, and probably most important, was Outlook integration. The big problem that we found with our customers, and generally with sales force automation system, is getting the sales reps to actually use the system and put data into the system.
With Fusion CRM, we have Outlook integration, which means a customer or a sales rep can update their customers, their contacts, their appointments in Outlook, and it gets synced automatically with the CRM system, that the sales rep is actually updating the CRM system really without having to enter into the CRM system, which helps a lot of that uptake and usage and gives the organization better data. Graco. Graco, the same type of thing, a standalone best-of-breed sales force automation system. They chose Fusion CRM for many of the same reasons. I call out Graco because this is a classic example of a rapid go-live from purchase to deployment, which in this case is just a SaaS pod. To go live was about 12 weeks for Graco to deploy their sales automation globally.
In ERP, the products that are included here, financials planning and budgeting, financials accounting, governance risk and compliance, a set of reports mainly around the close and financial reporting for SEC and other regulatory reportings, procurement and sourcing, costing, inventory, and finally, project and portfolio management. Here's a situation again where our breadth really matters. In the Red Robin case, Red Robin Restaurant, they were an existing Oracle customer. They were on JD Edwards, but JD Edwards World. This was the AS/400 solution before EnterpriseOne. In a sense, they had to go through a major transformation, particularly around inventory optimization within their restaurants. What was appealing to them was breadth of solution. They wanted to run it in the cloud. In a sense, Red Robin's an example of a customer need where they skipped a generation of application. They went directly from the AS/400.
They did not go to a E-Business Suite or any other kind of client server or internet-deployed mode, went directly to SaaS. The second key that was important here was standards for them, because they have integration to the inventory systems, to third-party barcode systems that they use within the restaurants to track inventory and track and report inventory. They used our open web services, the standard-based web services to inventory financials to do that rollout. They rolled out globally, not only ERP, but supply chain, as well as HCM in the cloud for Red Robin. These are live today. I called out each to try to give you a sample.
I showed you some coexist customers like Brocade that actually coexist in integrate, a coexist customer like Elizabeth Arden that kind of runs it more side by side, some standalone deals within CRM, like Graco and Key Energy, and then a more complete ERP representation like Red Robin. I call them out as a sample. These are the logos for the ERP and HCM customers that we've talked about at length at the show. Here are the logos for the sales and marketing, as well as service and support, so both the RightNow and the Fusion CRM customers. You'll notice from my presentation last year, the slides are getting more crowded and the logos are getting smaller, and that's kind of the sign of the progress that we talked about. Okay, why do we win?
It's reasonably simple to say depth and breadth, but I think one of the things that gets somewhat overlooked or maybe forgotten about is that in the different acquisitions that we've done over time and the teams that we've used to build Fusion applications, we have a tremendous depth of functional knowledge of people who've built these applications before. The team that built the Fusion Financials were the best-in-class developers from E-Business Suite, from PeopleSoft, as well as from Hyperion. Not only did the technology and things like social and BI allow us to do things differently, we also had a chance to take a brand-new look at what we built. Just some examples going forward. In ERP, we think we have the best-in-class GL from the start with PeopleSoft and E-Business Suite, as well as Hyperion.
Fusion Financials really combines those two like never before. When you post to the ledger, you now simultaneously post to the Hyperion consolidation queue. We took basically what was two best-in-class solutions, but had an issue because they were built by two different organizations, and used that functional domain expertise, which takes time to build within our teams from the experience of building these, and really fixed a lot of the business process issues we had with previous generations going forward. Just another reference for the Fusion ERP, Oracle, ourselves, so $30 billion-plus corporation global. We close our financials on Fusion Financials for Q1. We use it all three quarters and are live, and that's what we use to close our books in Q1. Examples like HCM in global.
If you recall back when we first built HCM, even PeopleSoft HCM and Oracle E-Business Suite HCM, those applications were not built as true global solutions. By that, I mean the following. If you had a global company, and you had people, a workforce that moved between countries, there were still definite issues within those systems on how you moved from one company in one country to another company in the other. Though you're moving between countries and maybe between legal entities, you want to keep things like salary, tenure, your vacation plan stays the same. All those things are measured. When you build the solutions at the first, wasn't really accounted for, that wasn't the way the world ran, we had a chance to rethink a lot of that, and have a lot of depth. In sales and marketing, I mentioned a couple of examples.
We really focused on having increased adoption, it really comes down to two areas. One is having it easier for the sales rep to get data into the system, where we have Outlook integration, as well as tablet integration on the iPad. Second is you had to make it more valuable for the sales rep to basically be able to sell more. When we had management tools and sales targeting tools like Sales Predictor and Territory Management, we were able to get better adoption. It'd give you a depth of functionality. Next thing is the breadth of functionality. I listed out the products.
In each product family, we think we have the biggest breadth of solution, so that wherever your problem lies, whether it was in talent management, core HR, payroll, distributed order orchestration, territory management, we think we have the right solution today and ability to expand. In addition, I said integration. We have those pieces integrated so that the customers we have today not only use our products, but are adding on sales and marketing to RightNow. Probably our most popular addition is the combination of Fusion HCM with Taleo. Sometimes it's starting with HCM and adding Taleo Recruiting, sometimes into existing customers that are using Taleo Recruiting in the cloud that now want to move to an integrated core HCM system.
The architecture allows a customer to not only have best-in-class integration, but because it's the same platform that you use and our platform as a service that Abhay talked about, if you want to extend the applications, you extend them in the same way that we built the applications going forward. It gives you very tight integration, gives you very good security, a nice user flow, and similar and the same look and feel. All combined with social, and frankly, some of the areas I think we'd argue that are just mandatory the way you do business for social today, no longer optional to add on or as a separate application. Okay. To summarize, what do we do? We want to deliver the world's leading technology and business applications to our customers anywhere. We have an extremely diverse existing customer base with therefore extremely diverse problem sets.
The way they approach those problem sets, either it's a point solution now where we compete best of breed, but we give those customers a path. In the case of Red Robin, we're doing a more wholesale replace of systems. We have that solution as well, whether it be in the cloud or in the private cloud, as we talked about before. Integrated from the start and integrated with social included, open standards for no lock-in, for ease of integration with third-party systems, and with a brand-new built from the ground up modern design. The common thread through all of our customer base, regardless of the heterogeneous problems I talked about, is speed and lowest cost. That's what we think we provide with the Fusion SaaS applications. Now I think I'm going to hand it over to Juan Loaiza, who will talk about Exadata.
I guess we're going to continue our whirlwind tour through technology. I'm going to give you a quick Exadata update in about the next 15 minutes. I'm going to go through some of what it is very quickly, then talk about our new generation technology that we just introduced, compare it to some of the other technologies that are in the market, talk about a couple of customers, and I'm quite certain that you'll hear more about Exadata later on from Larry. What do you think, Safra? You think that's a good bet? We're all very excited about Exadata all the way up and down the stack here at Oracle. Who am I? I'm Juan Loaiza. I've been working here at Oracle for 24 years on database development, working on mission-critical database capabilities.
One of the things that my group developed starting about eight or so years ago is Exadata, when we became frustrated with some of the limitations in the conventional hardware technologies. I have the same safe harbor, I have the same forward-looking statement. Let's start with some of the basic. Who's the target customer? Our target customer is the decision-makers for mission-critical databases. All the mission-critical databases in the world, whether they be OLTP systems or warehouses. Our available market is really that market, the market for mission-critical databases. That's what we target. Everything that's mission-critical database, we want running on Exadata. The big drivers are data growth, moving things online, the faster business decisions. That's the growth part of it. There's also a big transformation happening within companies where they want to simplify and reduce costs. That's the other part.
There's the growing part and there's the shrinking part. We're attacking both sides of it. Our strategy is pretty simple, which is we want to be the best platform for running any database workload, whether it's OLTP, data warehousing, applications, mixed workloads. We want to have the best performance and the lowest cost. We also want to be the best platform for clouds. What you'll see is over time, we're going to become, and we are becoming the corporate standard within many large corporations for their database platform. We sell the Exadata Database Machine, which comes in various different flavors. Our differentiation is we provide the best performance by combining leading-edge hardware with very sophisticated proprietary software. That's really where the magic is, in the very sophisticated proprietary software. We also provide a very complete solution.
Everything a customer needs, it's all pre-tested, pre-configured, it's ready to go. I'm going to lead you through a little bit of what is Exadata. As I mentioned, we're attacking the entire market. One of the things that's happening is people are claiming you have to specialize in a specific niche of the market to do well. We don't think so. We're going after the whole market, the data warehousing market, the OLTP market, the cloud market, the SaaS market, the applications market, the whole thing. There's big advantages in that because there's synergies among these markets. For example, the warehousing people want high availability and high security. The OLTP people want very high parallelism and very fast analytics. We don't believe these markets are really segmented. We think there's great synergies among these markets.
If we take a quick look at Exadata and say, "What is Exadata? What is this thing?" I talk about the leading-edge technology. What we have in there is everything you need to run an Oracle Database. What does that mean? It means database servers, it means storage servers, and it means networking in the middle. We started with a brand-new platform. It was first introduced four years ago. We've taken the most modern available architecture. What we're doing is scale-out database servers. Many of you that have followed the Oracle Database for a lot of years, you know that we made a big push towards scale-out database, starting about 10 years ago with our Real Application Clusters. That's been a big drive that we've had going on for a lot of years in database, and that's what we use inside Exadata.
We use scale-out database servers. The other big thing in Exadata is we did the same thing on the storage side. We're using scale-out storage servers. We don't use conventional storage arrays. We use lots of little servers, and we scale them out for storage. We're doing the exact same thing we did on the server side to the storage side, and we're going beyond that by putting database intelligence in the storage, and we're putting very large amounts of PCIe Flash in the storage. If you follow the networking industry, you know that the big trend in networking is unified networking. The same network for server-to-server and server-to-storage communication, and that's exactly what we do in Exadata. We're in the most modern kind of networking, unified networking, server-to-server, server-to-storage, interconnect. We're using the most advanced networking technology, which is InfiniBand.
We're kind of a step ahead there, too. That's just kind of a look under the hood and say, "What is in this thing?" The most modern server architecture, the most modern storage architecture, the most modern networking architecture all put together. That's what it is. The other big aspect of Exadata, it is a complete platform. That means it comes ready to go to a customer. It's delivered, ready to run. Everything is in there. Everything's been pre-tuned, pre-optimized, debugged. We've tested all the availability modes. There's also other advantages here. It is a complete platform. For example, we work with telecom customers that have very high-volume data. The expertise that we gain by working with them gets built back into this platform that's available for everybody. We work with financial customers on very high-availability systems.
The expertise, they put this thing through the wringer on all kinds of availability tests. Every time they find any kind of little issue, we work it back into the platform, everybody benefits. That's something we didn't have before when we just sold software, because when the financial guys found some kind of availability problem, they would fix it in their system, nobody else benefited. With Exadata, when the telecom guys find issues with volume, when the financial guys find issues with availability, when the military guys find issues with security, it all comes back into our platform. It all goes out in the market. There's a lot of synergies in having a unified platform. It's important to point out that, although I talk a lot about technology and a lot of cool new technology that's in Exadata, it runs the existing Oracle Database workload.
We run everything that runs on an Oracle Database, and it has run on the Oracle Database for the last 30 years. That includes SAP, E-Business Suite, PeopleSoft. These are the most sophisticated applications in the world. Those are at kind of a low level what's going on. We've also added a lot of special software technology, a lot of special sauce in Exadata, and I'm going to talk about three things here. One, I briefly mentioned, is scale-out storage. We scale out the storage. We also put database intelligence in the storage. Traditional system, there's a big computer over here where the database runs. There's a big storage array over here that contains the storage.
Every time we over here in the database want to access some data, we pull all that data out of the storage array, pull it into the computers, process it, throw it away, and the very next query has to do the exact same thing. That became the big bottleneck for database processing, is pulling all the data out of the storage arrays on every single operation, moving it over to the compute servers. The big kind of thing that we started with Exadata with was moving the database intelligence. Instead of moving a mountain of data to this little program over here, we moved the program to the data. We put database intelligence, we put a lot of processing power and database intelligence directly in our scale-out storage. That was the first big breakthrough in Exadata.
Another thing we've done, I'm going to talk more about the way we do memory and flash in Exadata, is we've put a lot of memory technology. That's the latest kind of generation of technologies in database. We're using what's called PCIe Flash. Flash technology has started to enter into the enterprise world. The way it's mostly entering now is what's called flash disk drives. What that is you take flash chips and you make them look like a disk drive so they can just slot into an existing system. Which is a great way to start, but the trouble is, making it look like a disk drive loses most of the performance benefits. The next generation of flash technology is PCIe Flash.
That's flash that fits directly into the motherboard of the servers, right next to the CPU with very fast access between the CPU and the flash. That's exactly what we use in Exadata. In a single rack of Exadata, we have 56 of these flash PCIe cards that give us extremely high performance. It's well beyond what can be achieved with a traditional flash drive technology. We've also written a lot of software to take advantage of that and make it transparent, put database intelligence into the flash. The third big thing I want to mention is our compression technology. Mark Hurd mentioned this. We have a specialized form of compression that we use in Exadata that we're able to do because we have a lot of processing power in storage, and that's what we call Hybrid Columnar Compression.
The benefit of that is for data warehouse workloads, we can achieve 10 times compression. Mark talked about that's a huge financial difference. For archiving workloads, we usually achieve 15 times compression. That is a gigantic amount of cost saving. If you can squeeze the storage down by a factor of 10 or a factor of 15, a customer gets a huge financial benefit. All these technologies work together. The fact that we're able to squeeze the storage means we also are able to fit a lot more data in flash. We're able to profit it in the storage with our processors in the storage. All these work together in a form of virtuous cycle. That's what Exadata is at the technology level. Let me talk about what we just introduced.
Larry introduced this on Sunday, which is our next-generation database machine, our fourth in the line, and it's called the Exadata X3. We started in 2008, primarily focusing on warehousing. In 2009, we expanded that with OLTP and affordable VLDB. We added a scale-up version of Exadata in 2010, and now we have our new X3. There's a lot of changes in X3, but the big headline change is it's really an in-memory machine. We've expanded the memory capabilities both in the hardware and the software, such that for almost all customers, they'll be running almost entirely out of memory now. Previous generation database machines were really primarily disk-based machines with a lot of memory to accelerate. What we really have now is primarily a memory-based machine with a lot of disk to expand the capacity of it.
The focus has changed, we've really turned the corner in the industry on disk-based versus memory-based. This is really the embodiment of that change. Part of the reason for that, which I'm going to mention, is we've quadrupled the amount of memory in X3. Typically, what you hear from year to year is an advancement of 20%, 30%. That's a very fast advancement. That's Moore's Law. That's what you typically see in our industry. Between our last year's version of Exadata, the one from four days ago and the one from now, we've quadrupled the amount of memory in the machine. It's 400% extra. That's a huge change, and that's really changing not just on a qualitative basis, but in a quantitative basis, how users are going to use this.
We have a massive amount of flash, and we've written a lot of software to make sure all IOs go to flash. That's the second part of it. Okay. Just drilling down into that a little bit more, what we have in Exadata, we're blending disk technology, flash technology, and DRAM technology. What we want to achieve is the best of all possible worlds. We want to achieve the cost of disk, the IOs of flash, and the speed of DRAM. We've written the software to make that happen. It's as affordable as disk, as fast as DRAM, with all the IOs you get from flash. It all happens automatically. That's what we call this mass memory hierarchy. That's the software that blends all these. The hardware has all the components, the software blends all these together.
If you combine that with our compression technologies, we can get up to 40 terabytes of data in DRAM, and we can get over 200 terabytes of data. Still have disk. We'll be able to keep backups on disk, historical data on disk, documents, images, things like multimedia files can go on disk. All the active business data, it will now really be in memory. That's the big thing that's happening in Exadata. Now, to put a little perspective on this, if you go back a few years, I'm going to compare the performance. Today, with this machine, we achieve 1.5 million IOs per second. Perspective, if we go back a few years, that would have required 150 storage array frames to achieve. That's what we achieve in a single Exadata rack. Same thing with data scan rates for warehousing.
You would have had to have a gigantic configuration, bigger than almost any customer in the world has, to achieve the performance that we achieve in a single rack of Exadata. Okay. It's a true scale-out architecture. We can achieve performance. A lot of people have flash. A lot of people have memory. What they don't have is the architecture to take advantage of it. This is just one little diagram that shows it. We can achieve 100 gigabytes per second from flash. Again, you guys don't know what that number is because you don't live these numbers like I do. A very large storage array, an enterprise storage array, can achieve typically somewhere around four, five, six, eight, maybe 10 gigabytes per second. Those cost millions of dollars.
The very biggest storage arrays you can find, the latest and greatest, the ones that came out this year, they claim they can achieve 50 gigabytes per second with the maximum possible configuration of a storage array. From the leading vendors, at the top end of their product, a single rack of Exadata achieves 100. There's flash and there's architected for flash, and it's two different things. We're very highly optimized, very highly ready for flash technology. Just to complete the technology thing, I talked a lot about memory. There's a lot of other changes in there. We've increased the amount of CPU, we've increased the amount of connectivity, we've decreased the power, and we've left the price the same. The Exadata X3 that we released last Sunday is the exact same price as the X2 that we had before.
We've not increased the list price at all, even though we've quadrupled the memory, increased the CPU, increased the connectivity, decreased the power usage, exactly the same list price. The other thing that we announced last week, this Sunday, seems like last week already, is we've increased the breadth of the product line. Exadata has come in a quarter, half, full, and multi-rack configuration. We introduced a new low end of the product line, which is the eighth rack, which is roughly half the cost of the quarter rack. That means we have a lower cost entry point, and this will primarily be used for smaller customers, smaller workloads, but also dev, test, and DR. We've had a lot of interest in that here at Oracle OpenWorld. It's actually going to be one of the most popular configurations.
It extends the reach of the product. Let me go over a couple of technology comparisons. I'll talk a lot, a couple of customers, and then I'll be done. This is a slide that we put together for Larry. He used this on Sunday. This is a competitor that we run into a lot, which is IBM. What this slide does is it compares the list price of comparable hardware. You take the main metrics, which are the number of CPUs, the amount of memory, the amount of disk, and the amount of flash. Those are the main metrics that you apply to a computer.
You say, "Okay, let's see what we have in a half rack of ours, and let's configure something that has the exact same metrics, the same number of CPUs, the same memory, the same disk, the same flash as what's in a half rack of Exadata with IBM technologies." You see there's a dramatic difference in just the hardware list price. If you just take the same bits and pieces and you put them together in an IBM config, you see there's a dramatic difference in list price. What it says on the upper right is, that doesn't consider all the extra technology that we have in Exadata. That's just kind of looking at the piece parts, not the architecture.
If you look at the piece parts, we're over 8 times lower cost, lower list price than IBM, and we have a lot of extra technology. We have huge amounts of CPU and storage. We have the ability to do the 10 times compression. We have InfiniBand. We have flash PCIe cards, not flash disks. We have a lot of extra technology, which gives us a lot better performance and a lot lower cost. It all starts at a much lower list price also. It's a very dramatic comparison, this is the actual list price comparison of one of the vendors that we meet the most in the market, which is the IBM Power series, their leading product with their leading-end storage array. A couple of slides on customers.
One thing I want to point out, we sell to a lot of different industries and a lot of different applications, and one of the big ones is applications. Our applications. Oracle E-Business Suite, PeopleSoft, Siebel. Here's a quick example of a customer who's a very large organic grocer that probably a lot of people here shop at. There's a typical example of what a customer like this does. They take a packaged application like PeopleSoft. They add lots of other stuff, like their ordering, their timekeeping, lots of other applications. They put it all in one Exadata. In this case, it's like a half rack with another half rack for DR and a second one for dev test.
It doesn't take a lot of hardware to run a very large and very well-known corporation running their business critical apps all the way across the board. They get the single vendor support, much better performance, much better availability. These are the benefits that we see. It's consolidating. It's combining all these different things into one platform, simplifying their IT, making it much lower cost. One other example, we are becoming the corporate standard at a lot of companies now. Exadata was introduced four years ago. We've gone through a process of establishing a presence in the market. What does that mean? First, people want to touch it, then they want to try it. They try one, right? They say, "Okay, we're going to deploy one. We're going to see how it works. We're going to benchmark it.
We're going to prove it." That's kind of the first stage of adoption in the market. Now we're moving to the next stage, which is, "All right, I have one, I have two. I really like this. I'm going to adopt this broadly within my corporation." I have a few examples here. For example, Procter & Gamble, we started with the trading there. We've now added shipment data. We've now added planning data, trade fund management. The footprint, the breadth of the applications within P&G is increasing all the time. U.S. Customs and Border is another great example. We started there in 2010. Now we run practically all the mission-critical applications. For example, when you come into the U.S. and you hand somebody a passport, that's now processed through Exadata.
All the cargo that comes in the U.S., everything that comes from China and everywhere else, has to be processed by U.S. Customs and Border. There's millions and millions of these every day. It's extremely mission-critical. It's all processed through Exadata. Again, the breadth, it started in one place, it's spreading throughout. That's an example of mostly an OLTP system. They have some warehouses also. Very mission-critical. If it stops working, the ports shut down. The airports shut down, the ports shut down. It has to work. It's extremely mission-critical. Another example here is Turkcell. They're spreading within their warehousing and analytics. Actually, I want to show that one interesting thing there is a petabyte of data. A petabyte is the next number you start hearing about now. A petabyte is 1,000 terabytes. It's that next higher value.
A petabyte of data in less than three racks of Exadata. That's the power of that compression technology that Mark talked about. It's a petabyte. For somebody who's been in this for over 20 years like I have, it's an amazing number. It's like fitting 1,000 people in a Volkswagen Beetle. We're squishing a lot of stuff in there. It's a petabyte of user data in less than three racks. It's just an amazing thing, and it's live. The use cases keep expanding. We started things with call detail records. Now we have subscriber data. It's just another example of you start with one, and then the footprint grows within the company. I'm going to wrap it up there.
What I wanted to conclude by saying is, our goal, we started with this four years ago, and remember I said the goal was the best platform for Oracle Database. Best for OLTP, best for warehousing, best for cloud, best for consolidation, best for apps. I think we've now achieved that goal. We are the best for all those things. We've achieved that goal, and we're just going to continue to innovate. There's a lot of stuff that we have cooking in the labs that's going to make it even better. I think I can confidently say, and I've talked to a lot of customers in the last week, and they will tell me we are the number 1 platform for OLTP, warehouse, consolidation, cloud, and apps right now. I think Hasan is next. Thank you.
Thanks, Juan. I'm here to talk about Exalogic. I manage the Fusion Middleware development in the Java development organization. What I was going to do is really focus on. I guess the standard disclaimers first. I was going to focus on Exalogic. I didn't want to repeat everything that Juan said, but Exadata is engineered system for database workloads. Exalogic is an engineered systems for applications, Java, middleware, other applications. The same principles around engineered to include hardware, software, the storage, the network, the compute. All the value in terms of engineered at Oracle, standard configurations, customers who use it, as we learn from it, we bake it all in. You get much faster time to value. All of those principles apply across all of our Exa line.
I'll focus on what. It's built to run Java, Oracle applications, other applications, non-Java applications, compute platform, across all your workloads. We offer physical and virtualized environments. We launched Exalogic, I'll talk about that. We started with physical. We offer now physical and virtual workloads. We provide a complete management framework to manage the entire infrastructure. On the middle-tier side, if you talk to customers, there's a big problem in terms of manageability across all these different. Particularly, that problem actually gets compounded. All the benefits of virtualization we understand, but one of the drawbacks is you have a lot of management complexity. We've built all of that into that complete solution. As I said, the architecture is the same across our Exa platform. InfiniBand network is a building block from a networking perspective.
We integrate with Exadata because as you can imagine, most of the applications obviously have a database behind them. One of our very common configurations is Exalogic and Exadata running together. The target customers, like I mentioned, Oracle application customers-based applications, another 120,000 bespoke ISV applications, Java and mainframe applications. We have over 340 ISVs which are certified on this platform. The key drivers, again, increasing demand, the retail utilities. These are still very, very performance sensitive. All the metrics you see around for each extra millisecond, how many people do you lose from your website? How many people put stuff in the cart and don't check out, et cetera. These are very intensive performance. At the middle-tier level as well, people want simplification in their environment. Again, Exalogic provides a great platform to do that.
Of course, all IT organizations, including everybody I talked to this week at OpenWorld, wants to see how it will save them cost within the first year. In terms of adoption triggers issues, that's still you'd be surprised even though you have middle-tier scale-out, a lot of applications still have performance implications. New projects, application upgrades, application life cycle includes hardware upgrades. Hardware at a certain point at which they do refreshes. Those are kinds of areas where we're seeing triggers for Exalogic adoption. As I mentioned, we've been two years in the market now. Very strong customer adoption across the globe, across all industries. You can see some of the name brand customers here. These customers in a little bit of detail. Ingersoll Rand, they were looking for improved customer experience.
In their case, this is another one of those Oracle application, E-Business Suite Release 12, Fusion Middleware underneath that. This is Exalogic and Exadata together. Six system, the IT spend, and a much faster response time. Just in terms of customer experience management, deliveries, quality of products. They, again, cost benefit, but as importantly, much better. Being on a one-time basis, ongoing reduction of 40%. Another example is Hyundai Kia Motor Company. Again, you're familiar with the company. In this case, they were looking at better scalability. This was an IBM. This is middleware workload, so content management, Oracle Content Management. In their case, again, higher throughput. They were automating document-related workflows. In this example, something that took them four and a half hours to do was done. Mostly driven by increased productivity, but also cost savings at the same time.
América Móvil, this is the fourth largest telecommunication company, mostly focused on Mexico, Latin America. This is half a rack of Exalogic. Really that performance requirements, in this case, 35 times the throughput. This was also interesting because they got it up and running within a month, and they saw ROI within three months. Again, that's a key point where you get the entire system. Just our customers, when you talk to, you'd be surprised at how many really sophisticated IT organizations actually value that aspect a lot, which is, even though I have the people and I can put the system together myself and I've optimized it, et cetera, but I would rather not have to spend money on that. If you've done the engineering to put it all together, that helps me get much faster ROI. Last response time, a big focus.
This is actually a lot of Oracle products running on this system, WebLogic Server, E-Business Suite, SOA Suite, some custom Java applications. This was a quarter rack of Exalogic and a half rack of Exadata. Again, cost, and performance. Again, improvement in terms of the resources, the human resources it took them to operate the system, and operational benefits. As you've heard, adding to existing hardware. The secret sauce is in how we engineer the products together. What are the couple of things I want to highlight in terms of the engineering benefits that we have? One thing is Exabus. I just want to spend a minute mentally, the InfiniBand networking that we have, while it's a faster network, higher throughput, lower latency, it's the question of how we re-engineer our products to take advantage of that.
In this case, for example, if you have regular networking, you have a lot of time. InfiniBand, we can do a much better job of transferring these messages, both in terms of the speed at which they go back and forth and also the reliability. We can take a Java workload and actually have it run much faster. The beauty here is that, while this directly to everything that's running on top of this, you don't have to change anything from an application perspective. This is standard Java application. All the smarts that we build in really get leveraged then by existing programs, existing applications. That's really the value here of not having to change your applications, run your existing applications, and you see the benefits. It's things like these that we've done that help you achieve that.
Another capability I want to highlight is the Oracle Traffic Director, which is really an application delivery controller. Being able to manage the workload in terms of incoming workload, service levels, is taking the software and integrating it much more closely with the hardware platform. In this case, processor-level hardware encryption, SSL security type configurations. Really not only provided this capability out of the box, we found in networking equipment. It's not just that we are putting that capability into the platform, it also gives application developers access to being able to control these, as opposed to having to call networking guys, which is always a big problem, present layers, to be able to get that application performance. This gives the control to the application developers, more importantly, really provides an additional capability built into the box directly.
We also announced the next generation of the hardware platform for X3. Mark isn't here yet, but I'd be happy to have him talk about it as opposed to Juan, who doesn't like him talking about X3. Exalogic X3 is the next generation of the Exalogic platform. Again, much more processing capacity, much more memory, flash, even better performance. Again, the same principle, it's the same price. This is 50%-100% more compute capacity now available, as of, X2 to X3. It's at the same price. In terms of performance, when we launched Exalogic, we talked about 2 to 5x better than the commodity hardware that's out there. For an OLTP application, for a Java application, that was a pretty big deal. As you saw in the slides earlier, we actually have seen much better performance in customer installations, actually, even beyond that.
These are some of the benchmarks that we look at, with the improvements in the software and the hardware this year, in terms of the Engineered Systems, we can actually now do another 3 to 5x jump over what we had last year. That's now cumulative. You can multiply that out. These are kind of the Java benchmarks that you see. A lot of our customers are using these for our applications. We've done benchmarking again in our labs as well as talking to customers and looking at customer benefits. This is again, Exalogic and Exadata combined. Like I said, most of these applications, all of these applications have a lot of application logic, but a lot of obviously database processing as well.
These are benchmarks that we've run across Oracle E-Business Suite, JD Edwards, PeopleSoft, Siebel, running Exalogic and Exadata in terms of both response time and scalability. Across both dimensions, much better performance, and again, we are seeing this in the customer base as well. This also applies to our vertical applications. These are some of the ones that I mentioned, which are very performance sensitive. Web commerce with ATG, retail, utilities, communications. Here, beyond just the cost and the management benefit, these are really top-line business benefit type applications. Having that extra performance actually is very important. Again, we are seeing fairly significant adoption in these segments, the retail, web commerce with ATG, across these. Again, we have the benchmarks to show the improvements again across scalability, response time, the typical metrics that customers look at. What are the competitive differentiators?
Just to walk you through that very quickly. Performance, like I said, breakthrough performance, Java, middleware, Oracle applications. We introduced virtualization. I just want to make a point about virtualization. Obviously, technology's been around for some time. Vendors have been doing it for a while. Most of the virtualization adoption in the IT space has been around development test-type systems. Mission-critical systems like Oracle E-Business Suite, like Siebel CRM, typically have not been virtualized because that technology is really not ready for those kinds of business-critical applications. What we've done, where we are differentiating ourselves, is bringing that virtualization benefits to mission-critical applications because we have reinvent virtual machine, all the low-level components that let us provide that level of service for these mission-critical applications. With virtualization, we have fundamentally addressed a space that has really not been addressed effectively before.
Of course, we have a lot of these applications, so we can not only do a better job, but we're obviously using it internally with our cloud, but also with our customers. I mentioned about Exabus and integrated traffic management. These are some of the key features in terms of how we deliver that extra performance. I think the manageability part is also important because we have a complete end-to-end manageability, specifically when it gets into virtualization. One of the challenges, like I mentioned, you take hundreds of servers and you boil them down because you have virtualization, but now you have thousands of virtual machines. Managing that across your entire enterprise is a big problem. We have, again, the tools to help you lower the bar in terms of manageability for these systems.
Just in summary, with the Engineered Systems strategy, we have machines for Oracle Database. We have Exalogic for the applications, Java, Oracle applications, ISV applications, mainframe applications. Two key value propositions: performance, very high performance, lower cost of ownership. The secret sauce is, again, in the way we bring the software and hardware together and the innovations we do in the software. Again, we're seeing very good adoption, both Exalogic as well as Exadata and Exalytics as an Engineered Systems approach within our customers, including replacing existing hardware and also using it within our applications in stall base. Thank you very much.
Tom. Sorry. Okay. Hi, I'm Thomas Kurian, I'm Executive Vice President of Product Development for Oracle. I run the software development group. We wanted to break up the process of having you guys see 16-slide presentations back-to-back. I'll take a few questions for about 10 minutes on the topics we've covered, cloud, Exadata, Exalogic, and then you'll have a break, and then we have more to come on Oracle Database 12, Exalytics, big data, and then obviously Mr. Ellison will be here at the end of the day. Please go ahead. Ken, are there mics being handed around or Okay, I think there's a question in the back.
I got a mic over here, though. Hey Thomas.
Okay. All right. Okay. If you could stand up because it's a little hard for me to see from up here. Yep, great.
Hey, I had a very simple question. Maybe I missed it, but will customers be able to buy Fusion Applications pre-configured, assembled on ExaMachines?
Yes. We have customers who use Fusion Applications. The database and the middle tiers come pre-configured using a virtual machine template. You can unzip the virtual machine and get up and running very quickly, and there are customers who've done that, and they come on Exadata and Exalogic. Okay, next question.
Go ahead.
Yes.
Trip Chaudhary with Global Equities Research. I had a question on platform as a service. If you look at the industry landscape, almost every company is struggling with it, salesforce.com, with Force.com and Heroku are struggling.
VMware with Cloud Foundry, even though it's open source and polyglot for almost a year now, it's failing to get launched. Same with Red Hat. OpenShift is dead on departure. If you look at OpenStack, it is all noise, no traction. What gives you confidence that your product will gain traction?
Why do we think that Oracle's platform as a service will gain traction? It's very simple. The companies who have attempted to get into the platform as a service business were never people who actually had a platform. If you look at Red Hat, it doesn't have a platform. If you look at VMware, it's an infrastructure provider. We offer one of the world's leading platforms with our database and WebLogic. The important message we gave across to our customers, and which our beta trial customers who have been using the platform as a service love about the Oracle platform as a service, is that it is the exact same database, and it is the exact same WebLogic server that you run on-premise. It is 100% the same. Same APIs, same access to SQL, PL/SQL, Java EE, even the same console that allows you to manage it.
They can move their applications very quickly into the cloud if they want to use this, or they can start in the cloud and move back on-premise. One of the customers, Siemens, there was a video about him, and he was so excited, he said, "It felt like the Oracle Database was sitting right next to me," except it was in the cloud. It's that excitement, I think, that gets the large number of our customers very excited about it because they adopted Oracle Database or WebLogic as a standard. They don't want to pick another platform when they go to the cloud. The cloud is a delivery model, and I think that's what's very different about what we're doing. Next question. Yes.
Hi, Thomas. John DiFucci from J.P. Morgan.
Yes.
What you're doing with Exadata is impressive for high-performance systems, even the previous version, but now what you're doing with the latest. It's interesting to see you come out with an eighth rack, which I assume is customer driven. I didn't hear anything mentioned about the Oracle Database Appliance, which I've heard customers talk about sort of some of their, I guess, lower-end Oracle Database workloads. If you could talk to us a little bit about that and what your plans are there.
Okay, I'll talk briefly. The eighth rack was introduced because customers, if you're running an ERP system, for example, or a mission-critical system, typically people system and disaster recovery, then they have a pre-production, and those are typically identically sized. They have a development test, a soft instance, and those, they don't want to have identical size as production. They wanted us to give them a solution that says. Prior to this, their model was they would try and use a development test system that was on non-Exadata hardware, on something else, because they wanted a more priced, a smaller configuration. We introduced the eighth rack, exactly as you said, business that I want the audience to understand is with Exadata, we don't just sell one Exadata machine. There are many opportunities for us to broaden into these customers.
There's a way for us to sell an eighth rack for test dev, so we consolidate more instances, and then we have the storage expansion racks if you want to buy more. Now, the Oracle Database Appliance was introduced. It's a very different market segment we're going after. What we're going after is within enterprise customers, as well as within small and medium businesses, there are customers who use the Oracle Database. They wanted a highly available Oracle Database set up very quickly and easily. We have a lot of customers now using it. There's about 700 customers using it as of the end of May, when we last counted. That business is growing nicely for. Giorgio Armani, for example, is one of the customers, the fashion house. They have a lot of small databases, 35 of them.
They got the Oracle Database up and running on a two-node rack cluster in six days, which, these things are on the order of a month to two months, six days to get your application up and running. We see these two products as very complementary, and one of the things we are starting to see, for example, with customers like Thomson Reuters and others is they buy many systems from us now. They buy a bunch of expansion racks. They're buying Database Appliance in certain departments. We'll talk a little bit about Big Data Appliance as well, and they can standardize more and more of their data management infrastructure on the appropriate technology from Oracle. Okay. Yes, next question.
Ryan from Barclays. Can you talk a little bit about your view on in-memory in the database and using DRAM, like one of your competitors is saying, as kind of the only source versus actually what you're doing, which is a nice hybrid of having the hot data and the cold data and the active data in one place? Especially customers are thinking about the cost of things, because if you think about data growth and trying to put everything into DRAM, it's going to be very expensive. Just kind of maybe contrast your approach here compared to what some of your competitors are doing. Thank you.
Okay. As Juan said, our general view is vast. Customers always have a way of finding more requirements than what fits within the scale of that capacity, right? Years and years ago, when people went to 48 GB of memory on a processor, people were, "Wow, that's so much data." Today that feels like a fraction of what people really want. Our view has always been that there are of data access, I/O rates, and costs. From our point of view, should not have to change in order to decide where data is put, whether it's in DRAM or flash or on disk. Our database manages it for you, and we move data transparently between these layers. We also give you the performance of DRAM, the I/Os of flash, and the cost of a blend of DRAM, flash, and disk.
We don't think anybody, no matter what they may say publicly, because we have the breadth of customers and understand the requirements, no one is going to be able to match the performance and cost that we offer. Period. Next question.
Hi. Back here, Kash Rangan of Merrill Lynch.
Yes.
Thank you, Thomas. Trying to shift the discussion to the Fusion Applications side.
Yes.
Certainly you launched it last year and nice references coming through. Where are we in the process of getting a few references for customers that have all modules of Fusion and are running the entire business financials, the CRM, and HCM, some of the edge applications you've chosen to focus. I'm curious, when are we going to get to the point where it's the entire suite is completely referenceable in production? Thank you.
Okay. It's a very good question. I want to just make sure. We introduced the first version of Fusion Applications general availability April 28th. We are very happy with the progress. To get to 400 customers in a year is really good progress. If you look at competitors, for example, many of them say that everybody's comparing different time periods. They say, "Oh, you guys were working on Fusion since 1992." Sure. But if you look at the effective rate, we're at 400. Workday's recent thing was they're 320. They've been in the market three years. We're in the market a year. We feel comfortable on how fast we're growing. With regard to your question on when do we get customers who are moving the entire to the equivalent of a global single instance of Fusion, I think that's what your question was, Kash.
Most of what we're seeing today is the buying for Fusion is happening department by department. That is a central thing that we see in SaaS, is the buying decision-maker is not necessarily the departmental decision-maker. The HR executive says, "I want HR SaaS." The financial person may say, "Hey, I still want to keep financials on my premise on a different product set." Part of what we're seeing is not that people find that Fusion is lacking maturity to run a global single instance, but the decision-making that's happening within customers. Our view is there are some customers who are buying more of the suite. When I say more of the suite, ERP, they've finished the ERP go live, they're happy with it. They're picking up HCM.
There's a number who have finished the HCM go live with Talent and are coming back for ERP. There's a couple of major financial institutions. There's a couple of major, let's call it hospitality and retail businesses. Towards the end of this year, we'll start having more of them go live with their second phase in November, December, early January. Yes.
Hey, Tom, this is Adam from Morgan Stanley. Maybe just a follow-up to what you just said. What are you seeing in terms of Fusion customers in adoption? Are they buying one module, two modules? What's a standard footprint or a standard value? Does that change at all if people are going with a cloud-based version or an on-premise version? Is one commitment typically larger than the other?
Okay. It's a good question. The primary purchasing decision is happening at the level of what we call a product pillar. A product pillar, there are four of them. ERP, which is financials, procurement, project management, and sourcing. That's one product pillar. HCM is typically a second, which is global HR, comp, benefits, payroll, and talent management. That's product pillar number 2. CRM, which is install base, sales, marketing, incentive comp, quota management, territory management. That's number 3. Number 4 is supply chain. Supply chain is typically product information master, costing, inventory, distributed order orchestration. Those are typically the buying pattern. There's not a lot of differences in whether people are buying on-premise versus SaaS. The decision to move to Fusion is typically designed around a collection of business processes that a specific line of business wants.
They're making that decision at the level of the granularity of that business process. SaaS is more a deployment option. Whether you run on-premise versus SaaS is more a question of who's operating it for you. We're not seeing a lot of differences based on if they were on-premise, they buy a bigger footprint versus in SaaS, they buy a smaller footprint. Okay?
Last question.
Thank you. Karl Keirstead at BMO Capital Markets. Thomas, I just wanted to ask you a question. I know Oracle's been on this journey to the cloud for a while, but it feels to us like the pace is accelerating of late. You've obviously acquired more aggressively. This OpenWorld is very much about the cloud. I'd just like to get your perspective on what you're hearing from your clients that's motivating Oracle to step it up in the last year. What's the change in tone that has caused you to get more aggressive? Is it big data like Mark Hurd mentioned earlier? Are you sensing a tipping point on the enterprise adoption of the cloud? Maybe a little perspective on the why now.
It's an interesting question. I'll give you two answers. First of all, none of the software we wrote for the cloud happened in the last year. None projects that we showed at this event that was organically developed happened since last, meaning started after last OpenWorld. These projects happened to be under development for many, many years. Most of them started at least since that sometimes the projects get done at the same time, and that's the way software development goes. You finish a bunch of projects, you get a bunch of customers, and then at OpenWorld, we feel comfortable the software's mature enough to open the doors up for a lot of people. I wouldn't interpret to mean last OpenWorld, we met a bunch of customers.
They said, "Hey, Oracle, we would like you guys to move to the cloud." We rushed off and built something, and 12 months later, voila, here it is. That's one reality. I will tell you, all of the platform long-term, but Platform as a Service started more than two and a half years ago. Okay? That was one. The second thing is we do believe that the line of business decision-maker is increasingly influential in the purchase of applications. Number 2 to us is a very fundamental change that organizations are going through, similar to if you put yourself back in e-commerce.
When the Internet first happened in 1998, most organizations said, "There's this e-commerce thing." Very quickly, they realized doing e-commerce by itself is a problem because the customers get upset, and they call your call center, and your call center guys doesn't know what's going on with e-commerce. They want to return something, and your order manager has never heard of them. Your order management system needs to change. You need to deliver something to them on a schedule, so you need a view of inventory. E-commerce, if you look back 10 years ago, people realized very quickly that e-commerce really needed a transformation of core systems. To us, social is the exact same thing. It's going to transform all the core systems in companies.
The reason we stepped in was we felt there was a maturation of technology that we could acquire to then integrate into our core systems. In response to your question, I would say one factor is that we have seen this cloud, the movement to software delivered over the internet browser, for a long time, and we've been working for it for a long time. The fact that all the technology happens to come together is just a reflection of the fact that we finished a bunch of software projects. On the acquisition side, we have felt that our engineers have done a fantastic job building a bunch of technology, and there's a few areas where we did not have organic development underway, and we felt we needed to make an acquisition to fill out the suite. It's as simple as that. Okay.
Ken, you'd like them to take a break for how long would you like them to take a break?
15 minutes.
Ladies and gentlemen, we'll take a 15-minute break. Refreshments are being served in the foyer. We'll see you back at 2:50.
Ladies and gentlemen, please take your seats. Our program is about to begin.
It's an energy source.
Doctor, it's spiking again.
Still believe in heroes.
Ladies and gentlemen, please welcome back to the stage, Ken Bond.
Thank you. We're not saying that necessarily that you will have those superhuman powers if you're using our products, but your customers, who knows? A couple quick housekeeping items. One, is on your desk, you've seen the schedule. Also on page two on a pink sheet is an evaluation tear sheet. Very much would appreciate your feedback on the day. I apologize, should've mentioned that to you earlier. Secondly, just as a courtesy of everybody in the room, you'll notice we've got these very big doors. They do swing and they let in a heck of a lot of light. I'd ask, if you need to come or go from the room, over here on my left, your right, there's a door. If you could please come in and go from there, that'd be very much appreciated. Let me turn it over to Andy Mendelsohn.
Okay, good afternoon, everybody. Let's talk about database. After that. After 30 years, database business is going strong. This is the data from Gartner Group, calendar year 2011 survey. They estimate the database market, SQL RDBMS market grew about 16% in 2011. Oracle, they're estimating, grew faster than the market, about 18%. We took market share, moved up to 48.8%. The top two competitors lost market share, according to Gartner. We like that. What's going on? Why is the database business still thriving? Obviously, we're in the information age. Everybody's getting more and more data, so it's sort of a natural growth. There are three big trends driving database. Big data is a very hyped up area right now, and it mostly means big data analytics. Analytics are traditionally done through databases and data warehouses and our BI tools.
All this big data commotion is really helping drive the database business. We'll talk a little bit more about that a little later. Clouds and consolidation are actually big drivers in our IT departments of the big enterprises. The story is pretty much the same. You got the flat budgets, but you have to do more to help the business grow. They're all looking to lower costs, and the key way to do that is around consolidation, and we'll talk a little bit about that. Engineered Systems, you've heard Juan earlier talk about Exadata, and we also have the Oracle Database Appliance Thomas mentioned. The combination of the database and Exadata is really powerful. Our competitors have nothing like it. IBM, Microsoft, nothing. Exadata becomes another differentiation for the database products and really, again, helps our go-to-market around database.
This is sort of our high-level strategy statement for the database. The one key thing that we've been focused on for a long time at Oracle is we build our core database product to deal with all the different workloads. Juan talked about this in the context of Exadata. We are trying to do transaction processing, data warehousing, consolidation, everything in between on our single Oracle Database software platform. That's how Juan delivers all those things on Exadata, is like the core database is designed to run really well for all those different workloads. Exadata, as Juan mentioned, the storage technologies were really limiting the ability of our software to really scale out to the biggest data warehouses. Exadata originally was designed to solve that problem.
Now that we have this scalable platform, our data warehousing business is doing much, much better, and as well as all the other businesses. We're going to big data. In big data, there is this new platform people all want to know about, called Hadoop. We've embraced that in the database group, and we are going to market, talking to customers about big data, and talking about our data and our Oracle Big Data Appliance, which runs the Cloudera Hadoop distribution. We are going to market in that space as well. We are building features like high availability, performance, scalability, to address the needs of all of our customers, whether they're in the BI space, data warehouse space, OLTP space, high availability, data optimization and security, et cetera. Finally, we have always evolved the database to take advantage of the latest innovations in the hardware space.
We continue to do that. The Database still works, of course, across all the different vendor hardware platform. Of course, we also have our own platform, Exadata, and our Database Appliance platforms. Again, we take advantage of every opportunity to leverage those technologies as well as we build out the Database. This is a graphical picture of all the different products we have. Bottom, or actually sort of on the side, we show the database machines. Thomas just talked a little bit about the Oracle Database Appliance, which is the one on the lower left. Of course, Exadata on the upper right. These are very complementary technologies. Oracle Database Appliance forms the low end of our database machine family.
It was originally thought to be targeted primarily for the SMB market, for customers who wanted high availability, but didn't have the expertise to deploy high-availability solutions easily. We're also finding that our enterprise customers see the Database Appliance as a nice complement to Exadata. They're going to use Exadata, and they're using data. One of the customers the other day at OpenWorld was telling me, "Database Appliance is really cool because I have places out somewhere in Asia, and all they've got is a closet, and Database Appliance is a perfect place to throw a database machine in a closet. I'm not going to maybe run Exadata." We're very excited about both of them. Then, of course, we have our traditional, our core Database family of products. We go to market around, we have the additions to the Database, Enterprise Edition, Standard Edition.
Then we have a set of add-on products, or the options that we sell as well on core Database product. These slides just talk about some of the key different areas we sell these options. These options tend to be very highly differentiated technologies. In performance, for example, our Real Application Clusters and our partitioning technology are there. Maximum availability, technologies which are also very highly differentiated, market-leading products. In the cloud space, we have the newest technology we just announced that we're going to deliver with 12c. I'll talk about a little later. Data optimization refers to things like our compression technology for lowering storage costs by compressing your data and still giving you really good performance even after it's compressed, which is not what most people expect. Development tools, we just give those away for free.
Then on the right side, of course, we have got things like security products as well, that we have some very interesting capabilities over there. I'll just go on from there. What we're going to do in the next few slides is just walk through a few customer success stories. I think one of the areas where I still hear, scale, you need some Hadoop thing or NoSQL thing or something. Juan pulled up the Turkcell petabyte data warehouse, so I thought I'll show you another one. This is SK Telecom, which is the South Korean telco. They are putting all of their billing information in. They are loading 10 terabytes a day. That's a lot of information. They have over a petabyte of user information. This slide is a little out of date.
They're actually now up to three Exadata full racks running this data warehouse. They're taking technology. They're getting 10 to 1 compression using our HCC compression on Exadata. The key story here is it's another example of a petabyte data warehouse in Oracle. This whole notion Oracle can't scale has gone away with the Exadata platform. One of the interesting things, if you refresh that hardware, virtually every one of them is going to Exadata. It's almost 100%. Okay. Of course, everybody knows Oracle is great for OLTP. We've been running all the big e-commerce sites, and huge banks use Oracle, and all of your companies probably use it. One up anyway. VocaLink is a payment processor in the U.K. They actually rolled in an Exadata system early 2011. They're doing 20 million transactions per day.
The thing that I thought was most impressive was the customer's quote, where he said, "This thing has had 100% uptime" the system in production in early 2011. That's pretty good. That's what you need for running these kind of payment processing systems. They're using our maximum availability architecture. They're using Real Application Clusters and Active Data Guard to provide an active-standby. Again, they're using a lot of our highly differentiated technologies for running the system. Okay. The next few systems I thought I'd go into is consolidation. As I said, all the IT organizations are trying to lower costs and trying to do it through consolidation, and go through here. Garmin, you all know Garmin. They build the navigation systems. They help you navigate in airplanes and boats and cars and things.
They did a consolidation exercise, you can see they consolidated things like their packaged apps, E-Business Suites, Supply Chain, Hyperion from Oracle, and some. They moved them from 12 separate servers onto a single Exadata V2 box. They're taking advantage of our Data Guard technology to make it highly available, so they have a primary Exadata half rack and a standby half rack as well. Some of the metrics, certain times better performance, great uptime, 99.95%, and they are in position to grow. Exadata is a scalable platform. They need more capacity, they can just add it on-demand as they need it. Okay. This is another nice one. Company, they're a big IBM mainframe customer, they're a SAP customer, and an Oracle customer. Again, they decided to do some consolidation to lower costs.
The first thing they did is they consolidated all their SAP apps on the Exadata box. The next thing they did is they consolidated a bunch of mainframe. Look at the variety of the kinds of applications people are consolidating. It's like, there's sales and logistics and master data management. It's a combination of things that are closer to OLTP, and some are more like BI applications and, of course, SAP. You really need a platform for consolidation that can run all the different work. I guess my last little comment here. Well, on the SVP side, one of the cool things is we got 50% compression of the data over there, up to 100 times better performance. They're saving $125K per year in operational costs. About, ask Thomas about, SAP and HANA.
This is where I get to say, "Well, why don't you go ask SAP about when HANA will be mature enough and functional enough to have the HA technologies and security technologies and everything else you need for doing consolidations like anything?" Obviously, this customer decided it wasn't going to be anytime soon. Okay. Finally, Database as a Service. Right now, it used to be everybody talked about consolidation. Now they're sort of moving to Database as a Service, which is sort of the next phase of consolidation strategies. is the big pioneer in this space.
They've been doing Database as a Service, where basically all the different lines of business in their enterprise go to the central IT group, and they say, "I need a new database," and they get it provisioned rapidly on top of a centralized infrastructure, provide this on the cloud. A lot of customers want to do this on their private clouds. Commonwealth, as you can see here, evolved. They were always using Oracle Real Application Clusters for doing this. They first started on big Unix servers. They moved to commodity Linux server. We can see the cost, the return on investment, 150% over five years. As I said, this has been a long-term operation they've been doing, 50%-60% reduction in operating expenses. This is what every enterprise wants to do. This is the model Commonwealth Bank is really blazing the trail on this. Okay.
That's a good segue to what's going on in 12c. 12c is the first major release of the database since 11gR2 that shipped in 2009. We've been working on this for a long time. Some projects have been going on over five years. Major, major release, 3 million lines of new code, 15 million hours of testing. We're now in beta 2. We went into beta early February. The beta 2 release is now out there. Oracle is actually running some production systems. [Inaudible] Even know that they're using 12c today. Their favorite system called Aria, which is the system everybody uses all the time at Oracle for looking up the names and phone numbers of people and shows you the org charts and whatnot. That's production on 12c. Okay.
To go through all this today, but just as for you to see the kind of areas that we focus on in a release from application development to big data consolidation, data optimization and compression, data warehousing, HA, in-memory technology. I'll do one feature, which is our favorite pluggable database feature. If you saw Larry on Sunday night, he talked a lot about this as the key technology for multi-tenant databases on the cloud. It turns out this feature was actually originally designed- consolidation on their own private clouds or do Database as a Service. It turns out the exact same infrastructure works fine on either a private cloud or a public cloud. The idea is you take, instead of having separate databases running on a server and using that as a way of doing consolidation, instead you database.
Now instead of having 10, 20, whatever the number of separate databases you have to manage, now you have one database to manage. So the big thing that customers want for lowering cost of ownership around consolidation of databases is lowering the labor cost. This feature, it has this attribute that we call Manage Many as one. So you plug in 100 databases into this one container, and now you can do things across them, like back them all up with one push of a button. Instead of backing up 100 databases, you now have backed up by pushing the button once, you've backed. You want to do a patch, you patch all the container once, and that deals with patching all the pluggable databases inside it. So it is really nice Manage Many as one capabilities.
It's got isolation capabilities so that all these different pluggable databases running different work. With things like that. So you can think of this technology as sort of a way of virtualizing a database. So just like VMs virtualize servers, this capability virtualizes databases. If you talk to our customers, you will find every one of them wants to deploy this as soon as they can. Okay. So let's see. With that, I'll just close, with what's great about the Oracle Database. The number one thing to emphasize is we are the only vendor to this day who really concentrates on building Stability characteristics for all the different database workloads, and that's the foundation for why Exadata is so powerful and can also run all the different workloads.
We continue, of course, to run across all the different OSs and platforms, but we also now are delivering our own Engineered Systems as well. Cloud consolidation is a key area where Oracle has highly differentiated technologies to make all that possible. Things like Real Application Clusters and automatic storage management and Exadata scalable storage servers are really critical for building out clouds. They're highly scalable technologies. Management and the new pluggable database feature for consolidate [Inaudible]. Business intelligence and analytics, big data is another area we're very strong in. Oracle as a company is the only one that can really deliver a complete platform from the Big Data Appliance for running Hadoop, to Exadata for running your big data warehouses, to things like Exalytics, that Balaji will talk about next for running all your BI applications. So a very strong BI offering.
Finally, our, what we call Maximum Availability Architecture is, again, a highly differentiated set of technologies for making your systems highly available, including things like, again, Real Application Clusters, Active Data Guard, GoldenGate are examples of some of our products in that space. Just to close, the summary is the Oracle Database, the RDBMS market and database business continues to grow. It's a major part of, obviously, Oracle's business. We're the leader because we're the technology leader. Our development organization is working hard to make sure we maintain our years of leadership in that business. We have a big, loyal set of customers who are all eager to uptake our 12c release as soon as And use things like pluggable databases and a whole lot of other innovations in that release. With that, I'll close. Thank you. Bye.
Good afternoon, everyone. Earlier, we were talking about data and how to make sure that the things like Oracle Exadata, Oracle Big Data Appliance, et cetera, bring the management and the enterprise-grade architecture. Business analytics is all about unlocking the value from it. If you look at Oracle's business analytics strategy, it's very comprehensive, yet it is very simple to articulate. First of all, we're talking about analytics across any data and any source. What we mean by that is really data could be structured data, it could be unstructured data, it could be semi-structured data. You want to be able to analyze this data together, contextualized and correlated so that you can get the integrated insight as opposed to a siloed insight. The second thing is any source.
It could be an OLTP system, it could be a bespoke application, it could be a Hadoop or a big data type source. Again, it should be able to handle any of those sources. That's a very important foundational principle to Oracle's business analytics strategy. The second thing is, once you have an ability to analyze this information, the type of the inquiry that could be various different types of tools and techniques. There is no one-size-fits-all. You want to be able to report on your operational data. You want to be able to do ad hoc analysis and query on both the operational and the historical data. You want to be able to do multi-dimensional analysis to slice and dice and get better insights.
You want to be able to connect it to the unstructured data that's coming from the social and other places so that you get a more comprehensive view. For that matter, you also want to be able to then create some kind of predictive models and other kinds of things to be able to predict future outcomes. Really the important thing here is when it comes to Oracle business analytics, it's about really supporting the full range of analytics as opposed to saying it's a one single tool would fit all the needs. It's also about performance and the TCO within that context. The other unique thing about Oracle business analytics is locations that we provide out of the box, and it's very comprehensive. There are 2 categories of applications we provide. One is Oracle Business Intelligence Applications or OBIA.
These are the applications really targeted for specific subject areas, sales analytics, marketing analytics, financial analytics, procurement analytics, supply chain analytics, where we actually know the sources. We have a very good set of repeatable use cases. We can, in fact, create and deliver a set of applications out of the box and then patch them, lifecycle manage them together. The second set of applications is what we call performance management applications. Alternatively, they're known as Oracle Hyperion. This is actually targeted for the office of CFO. They're actually focused around the statutory and the strategic business processes that the finance operations actually deal with at any given time.
They're around planning, budget consolidation, cost and profitability management, tax provisioning, a number of different things that actually are very critical and very complementary to the business process of the transactional applications that finance operations use in terms of the GL and accounts payable, AR, et cetera. Oracle's business analytics is very comprehensive. The other thing that's very important that's actually the slide is not talking about is the delivery mechanism associated with it. It could be on-premise, it could be on the cloud. That's from the delivery perspective. From a consumption perspective, it could be desktop or a laptop, or it could be on the mobile device. In fact, by 2015, Gartner actually predicts that about 50% of the analytic consumption is going to be through a mobile device and not through the traditional device. That becomes very important to us.
The one that actually comprehensively surrounds this tree, very similar to what we do on the data management side, and this is where the Oracle Exalytics really comes in. Exalytics is truly an engineered system in the sense that we actually have taken a hardware configuration that is really appropriate and right for the kind of the analytic workloads that we have. At the same time, we wanted to make sure that it is available at the best cost performance. In other words, best performance, but at a cost that actually makes best sense. Let's look under the hood quickly as to what is in the Exalytics. One of the most distinguishing features in Exalytics is the amount of memory, the terabyte of the DRAM that's actually available in Exalytics.
What this means is, combine this with the compression of about 5 to 10x that we provide through Oracle TimesTen, you can bring large amounts of data into the memory, then you can start analyzing this literally at the speed of thought. It means a lot of different things. It really means you can actually do very fast interactive analysis. You can ask the questions, you can get the answers very quickly. That means the quality of your decision-making suddenly is based on the kind of the detailed analytical set that normally you would be settling for. In the planning side, the same thing. You could run a lot more models. You can do more sophisticated cubes.
You can do a lot of different things that, in turn, give you an ability to forecast and predict what you want to do as it relates to the business and how you want to run it. The other thing that a lot of times people ask is, what's actually inside from a software standpoint? It obviously includes Oracle BI and Oracle Essbase, I'll talk very briefly about that. It also includes Oracle Endeca. This is an acquisition that Oracle made about a little under a year ago, is because, as we said, any data, any source means it could be unstructured, semi-structured data as well. Endeca is excellent at handling the unstructured data from an analytics perspective, we wanted to make sure that all of this actually comes together under one single engineered system. The other thing people also ask a la carte.
While it could be possible, one of the key things to understand here is that the power of the Exalytics really comes in when you actually bring all these different full range of analytics together. When you start asking the questions, when you start actually doing reporting on the operational data, very quickly you start asking, "Now look at my trends based on the historical data using the same context." From there, can I actually move and do some kind of a multidimensional analysis using the same context? From there, can I actually connect to unstructured data using the same context, and so on. That's one of the reasons why this real foundation where all these things really come together. Okay?
Exalytics is really about bringing all these things together and doing so at the best cost performance, then doing so obviously in a very much a speed of thought analytics basis. The market opportunities for Exalytics is, as we've been looking at it, certainly we expected it to be very large, and we've been in the market for about a year, and we're suddenly seeing that opportunity is bearing fruition quite a bit. The reason is as follows. There are three a lot of sense. One is there is a large enterprise data warehouse market where analytics is one of the fundamental application that sits in front of it. As the data warehouses grow larger, as the need for something like Exalytics really becomes much bigger. Okay? That's a big area.
The second area is Oracle has these applications in analytics space that I talked about earlier, all these applications can truly leverage the power of the Exalytics in customer base that we have, which is approximately 15,000+, who all in fact can use today the power of the Exalytics to do the enterprise performance management that they're actually doing from an Office of CFO perspective. The market opportunity is very big and settling yet as well. Very quickly, I don't want to spend a lot of time in the interest of time here, I want to talk to each one of these analytical tools very quickly. One is on the reporting and analysis, which is the traditional. People always ask us, how are you different from someone like SAP HANA? I think it's a very important question.
I think we have a fundamentally different philosophy. SAP says, in order for you to do any type of analysis in memory, you have to bring all the data into the memory. When you know that the data is large, particularly because you want the quality of your analysis to be based on the large sets of the data, as opposed to a sample set of the data, you do have a finite capacity in terms of your memory. If you ask SAP, what do you do? They just say, "Have 512 GB, add another 512 GB." You say, "Well, what if I exceed the TB?" Add another 512 GB, and so on. Suddenly, it becomes extremely cost prohibitive to bring all that. We think there is a better approach in the Exadata.
To be able to understand at any given time what's the hot data based on the heuristics that we actually have a pretty good strong IP on, that in turn allows us to really identify where the hot data is and make sure that that data is always available in memory. If the memory capacity is such that the data is actually larger than the memory capacity, then we use these heuristic algorithms to actually manage that process.
Thereby, we really do not require somebody to go in and say, "Just keep adding more memory." I think it actually works very well, in fact, many of the customers that are actually using SAP beyond a departmental application and beyond the single application, when they start actually looking at it from the standpoint of more of an enterprise-wide, they start realizing the same thing, and then they actually now. Very, very important thing to actually understand in terms of the difference between Oracle's approach versus the SAP and the HANA approach. Very briefly on the Essbase and the planning and the profitability side. The thing that makes me really talk about this performance management applications in planning, budgeting, et cetera. They also have in-memory, but they don't actually work together. It actually has to be rewritten to take advantage of HANA.
Hyperion Planning, every Hyperion plan can literally, all they have to do is point the instance of the Essbase they're using to the instance of the Essbase that's actually running on Exalytics, and then they actually can get the full benefits of Exalytics. SAP has to rewrite BPC, and they're not done with it yet. IBM has TM1, IBM has Cognos Planning. They're two different systems. I think it's a very important thing to understand the power of what we're really trying to do here, and that is to maintain this compatibility. Finally on the Endeca, this is where the big data fit really comes in. One of the benefits of the Endeca is it actually extends the structured analytics.
A lot of the unstructured is actually the sources is actually right now from big data, from the social and from the log files and other places. To be able to really handle both types of this through something like an Endeca, to be able to do that not based strictly on a known set of questions like, here is what I'm looking at, but to go into more of the discovery and exploration becomes very important. The reason is, in the big data world, you don't actually know what you're looking for. You think you have a lot of good information out in the social world that you're actually bringing it in. You also know there's a lot of noise there, and you want to signal from the noise.
In order for you to be able to do that, you start wanting to explore this data, discover this, and discover those patterns and make sure that you're really identifying that subset of the data that, in fact, is the data that's actually something you want to further analyze. That's really where the Endeca really comes in. That's why they have anything similar to that. Finally, as I said, there are about 80+ applications today that are certified on this platform, pretty much from the time we actually ship this. The reason is, while we have done a lot of things from an optimization standpoint for in-memory that we have done and the IP that we added, what we fundamentally felt was that the applications do not have to be rewritten.
That's why we kept the APIs fully compatible, and thereby, actually, many of our applications today run on this box pretty much from day one. The customer uptake has been phenomenal, as I said, but even me, from my vantage point, is the kind of the benchmarking that they do now. Almost every customer, either they do that in the context of a proof of concept, or they do it in the context of the actual production deployment. They're actually now starting to share with us many of their benchmarks. Initially, we saw we can, from a BI standpoint, anywhere between 6X-20X. As you can see in these examples, 6X-20X is actually the low end of what we were actually looking at. You have customers such as Pirelli, 60X faster, Verizon, 60X-300X. 80X in terms of the calculations for planning.
You have Pernod Ricard, as it relates to what they're trying to do from a finance operations perspective. Department of Homeland Security, something that I think we can all relate to, passenger analytics. In fact, they do passenger analytics and much more than what they're exactly doing. You can see the kind of the things they're doing, 20X faster slice and dice with the Exalytics, with the Exadata in the back end, with almost 2.4 billion rows. Doing that with 89 seconds versus the 31 minutes is what on average. In Midwest, a mid-size company, they are actually using this both for the BI as well as from a Hyperion perspective and 100 installations. You have a large company, world's largest. He says they try to do a trial close, they attempted to do it on a daily basis.
The problem was they were actually taking some of the recomputation because of the kind of the complexity they have, up to 24 hours. Sometimes, they can literally do it less than four hours. For them to realize their daily close vision is extremely possible now with the advent of Exalytics that they never had before. We certainly also look at the talks about in-memory. We definitely could not agree with that more. I think the important thing is to ask, is it comprehensive? Is it fully understood? Is there real IP behind it, or is it just throwing more memory at it and then making it cost prohibitive? This is query optimizations. You name it. You look at these requirements, I think you'll start finding that we're certainly breaking away from the pack here.
I think in summary, from an Exalytics perspective, we see a very large market opportunity for this. Because of what we have here from a value proposition standpoint, it's actually built on a business analytics strategy that's very comprehensive, very broad, and deep. It's very complete, and this is a key point. It's not about range of analysis, it's about doing it in an integrated fashion. That actually differentiates, again, us from others. Certainly, we believe as a result, it solves problems a lot better than where the competition is. I think the customers are actually seeing that, then they're suddenly paying it by the pocketbook. Thank you very much. Larry should be here around 4:00 P.M., I'm going to try and get through big data and then take a few questions. We'll follow up broader broadcast on big data.
I wanted to just talk briefly about big data to clarify some confusion on what big data is and how we're approaching the market with it, okay? What is big data? Every company, a big data company, many are trying to be both, right? Big data is to us, for people who are technical, it's characterized by four things: volume, velocity, variety, and value. Volume is large volume of information. Storage. Velocity, the data comes at you very fast, so you need to be able to rapidly ingest it and persist it on low-cost storage. These two are important things. Variety. They have very flexible data shapes. Like a weblog does not have structure to it. It's a giant record of a binary stream. Even if you parse it can have a variety of shapes.
You need highly flexible storage in order to represent that. That's very different than a traditional relational storage requirement that has, Data set has traditionally got very low-density information, okay? I'll give you an example. If an order entry system, if you lose an order, people are worried about it. In a weblog analysis or a social media feed, suppose you had 100 web servers and you lost an hour's worth of data. You don't worry about it because you're looking at aggregate patterns across it. Okay, it's volume, velocity, variety, and value. The reason I mention this is later on I'll explain why the variety and value has a certain requirement on infrastructure and why that's different from, there's a lot of confusion in how does that relate to relational databases.
Okay, the obvious thing about big data is it's an analytics opportunity. There's analytics in three ways. I got a lot more data, so I can analyze it. I can do a lot better analytics if I can access all of the data, because then I can look in much more fine grain. Obviously, because it's so much data, my traditional approaches to analytics may not be sufficient to go with it. Most people talk about the five things you need to do with big data. You got to acquire it, you got to store it, you got to process it, you got to organize it, and you got to analyze it. Okay, how does that work today? If you look at it, here's what most companies and most users are looking for.
I want to acquire it. There's a variety of sources and data types. It could be genetic data, it could be weblogs, it could be social media feeds, semi-structured and unstructured. Okay? First step is you need a collection of connectors to basically receive and suck in this data, then store it somewhere. When you store it, you can store it as raw files in something like HDFS, Hadoop file system, distributed file system. If you've got columnar shape to the data, you can store it in something like HBase. If it's key value shape, you can use something like NoSQL. If you look at it, there's already three types of storage right there. Files, columnar, and then key value.
Once you store it, you want to process it. This is where the technology called MapReduce comes in. It's basically a parallel infrastructure to run jobs to process the data. When you process the data, you need to extract metadata to understand what's that data. There's something called HCatalog that provides you the metadata about the data in the system. Increasingly, people want real-time event processing because they don't want all the data to just go in a batch fashion. Okay? You need to organize it, because eventually you need to do analytics. For organization, people may use an ETL tool like Sqoop, or they may script against it. Sometimes, if you're looking at raw textual streams, like the Twitter data or Facebook data, for example, you need text analytics in order to be able to parse it.
On the right side come all the different types of analytic styles of analysis people want. Graph analytics is, say, who's a friend of a friend of a friend on Twitter or Facebook. Okay? It's traversing a graph. Statistical modeling is to look at sampling and looking at key statistics and standard deviation. Data mining, spatial analytics are also different kinds of analytic types. Decisioning is saying, "If I got that data in my HDFS system, how can I run a decisioning algorithm to figure out what to present as the next recommendation for somebody to see a product or service?" Query and analysis is traditional BI. Information discovery and text analytics are obvious things that you guys know about. Big data and Hadoop. This is where most of the confusion comes from.
What Hadoop does, it's a parallel compute framework and distributed file system. Okay? At the core of it sits a distributed file system called HDFS. On top of that sits a parallel compute infrastructure called MapReduce. Hadoop's value is it provides a highly scalable infrastructure for managing data that has got great variety and low information density. Okay? Many people confuse this third point. Hadoop is a core element of Oracle's big data strategy. We use Cloudera as our Hadoop distribution. We have a very good partnership with them. Hadoop alone is not sufficient for big data. This is the central point I want to get across. It's massively scalable, but it's batch only. Okay? It does not allow you to do decision support or interactive analysis. Why is that? It's very simple.
To do interactive analysis, when you throw a query at an Oracle Database, we have something called an optimizer that says, "I'm going to calculate how long this query's going to run." The way it calculates that is by looking at where is all that data on blocks and disks, and roughly how long is it going to take me to retrieve all these blocks and then run the calculation and return the query. In Hadoop, the strength of Hadoop is that it gives storage, because it has to do that given that it's storing highly flexible data types. In giving you that flexible storage, it cannot compute very efficiently how long it's going to take to run that query. It's a fundamental mismatch to say why can't Hadoop do everything a relational database can do, and why doesn't a relational database do everything Hadoop does?
It's completely apples and oranges. Hadoop's a fantastic technology to store and manage flexible data types. In the act of storing that flexible data, it is impossible to perform the same kind of query optimization that a relational database does, period. It's not an issue of can it do it today. It is not going to be able to do it because the flexibility of the storage runs completely counter to being able to do a totally optimized query calculation in a fixed period of time. I hope that's crystal clear. What a lot of the leading companies, the Social Network, as you call it, that company, what they do is they store the unstructured data in Hadoop, and they run pre-processing and they add structure to it. The more structure they add to it, the more they put into a relational database.
Eventually they're running distributed batch calculations on Hadoop, but they want to be able to do interactive analysis in a relational database. The central thing I hope this gets across is three points. One, Hadoop is a good solution to store and manage flexible data, but by definition, it is not a replacement for a relational database. Number 2, the more data that goes in Hadoop, typically we see a 10X. If you put in 10 petabytes in Hadoop, you'll put a petabyte into a relational database like Oracle. The more data that goes into Hadoop, the bigger market that it opens up for Oracle. The third, people are not taking structured data out of Oracle and putting it into Hadoop. I'm sure if you went out on the web, you could find one strange person who has done it.
Typically, they're not doing that because you're using our Oracle Database to handle highly dense, high-value structured data, and it's the diametrically opposite requirement to extract the data from it and put it into Hadoop. Is that clear? I hope those three points are crystal clear. What a lot of customers are seeing is, therefore, Hadoop is a component of a bigger warehousing and analytics strategy, and that's where we're going. Oracle is offering the industry's broadest and deepest analytics strategy. What we'd like to do is to go through this in a lot more detail with you guys. Ellison is here, I'm going to hand over to him, then Ken Bond and I will schedule a follow-up call, where we can do a broader briefing on big data. Larry?
Thank you. Let's see. Is this on? It's on. Okay. I'll make my first request, which is turn the lights down, please. That's much better. Whatever you guys want to talk about, I'll do my best to address the issues and answer the questions you have. Okay. Thank you very much. This has been great.
Right here.
Oh, hey, Larry. Rick Sherlund. Larry, I wonder-
I knew that, Rick. I knew that.
Thanks. Larry, it almost seems again like it doesn't matter what the question, I know what the answer's going to be. It's going to be about the cloud.
It's all about the cloud.
Right.
It's all about the cloud.
It actually seems very interesting that I think for Oracle, that a couple of years ago, it didn't seem like you kind of got the cloud.
Let me stop you right there. I think it's very important. I want to really admit, even though I founded NetSuite, the first cloud company, I absolutely didn't know it was a cloud company. The first thing, let me be clear, I'm not an investor in NetSuite. I own it. I started it. It was my idea. Be very clear. I called Evan Goldberg, and after mBed failed, I said, "Evan, this time we're going to do what I want to do." Evan started a graphics company first. I wanted to do accounting on the internet. It was my idea. It was 1998. Six months later, Marc Benioff found out what we were doing, and he came up with salesforce.com. He said, "Well, if you're going to do accounting on the internet, we're going to do sales force automation on the internet." I don't get the cloud.
The first modern cloud company was my idea. I didn't sell any of my stock. I still own it all today. When Marc said he wanted to do sales automation in the quote, of course, we didn't. The problem was we just didn't know it was the cloud. We thought it was accounting on the internet, and we thought it was sales force automation on the internet, and eventually, we thought it was SaaS, so we thought it was software as a service. We just didn't know it was the cloud. What I objected to when I said suddenly the term cloud became popular, because in 2006, amazon.com rolled out EC2, the elastic compute cloud. The term, the very charismatic term cloud, was retroactively applied to NetSuite and salesforce.com and all these other things. I kind of ridicule out.
Everything's cloud, not just Amazon, but Salesforce, NetSuite, everybody's cloud. I made fun of the term cloud. As far as not getting the cloud, I'm pretty sure the first modern cloud company was entirely my idea. Go out and do some fact checks. Now, if anyone, I'd love to hear anyone refute that. It's 1998, you can check it, before salesforce.com come up with a modern cloud company. It began then in the late '90s. Now, this idea I didn't get it is interesting. I think I invented it. That's a pretty big separation. I don't think Edison got the light bulb. Like he just didn't get it. In 2004, I decided we should rewrite all of our applications, and they all should be SaaS applications. It was eight years ago. We had to rewrite everything. We have a huge inventory of applications.
We had to take hundreds and hundreds of people off what they were doing and putting them on writing these Fusion applications. It was eight years ago. Maybe I didn't get it then either. I used to go around talking about this thing called the network computer. Now, the idea of the network computer was you'd have a simple appliance to access the internet. You'd have a simple browser on the network computer. That was my idea, too. The network computer, remember that? Just take a simple browser, and all the complexity would be hidden from the user on the servers on the internet. I just didn't know it was the cloud. I just didn't know the word cloud. I don't accept the notion I didn't get the cloud. You got a second half to your question.
Yes. I was wondering if you could just talk about your approach. I didn't mean that you didn't get the cloud. Can you talk about your approach to the cloud? Multi-tenancy is not something that you historically have embraced with Fusion.
Let me be clear. I didn't embrace multi-tenancy built into applications. I didn't think you should at the application layer. The cloud has three layers. SaaS, people like Salesforce, NetSuite, Workday, they write applications, then they want to be able to run lots of different customers efficiently on a minimum amount of hardware. Because they're application companies, they build multi-tenancy into their applications at the application layer. I think that's a horrendously bad idea, but you had no choice in the late 1990s. You had no choice. There are two other places you can implement multi-tenancy. The second layer of the cloud is PaaS, the platform, database, programming languages, middleware, business intelligence, those kind of things. We think the right place to implement multi-tenancy, in other words, putting a lot of customers. The right place to implement multi-tenancy is not inside of the application.
Instead, it is inside of the platform, specifically the database. The database should be responsible for managing data of different customers. Long time. Took us almost five years to re-engineer the Oracle Database 12c to implement multi-tenancy at the database level in the database code, rather than the application code. Salesforce does it at the application. We do multi-tenancy at the database level. 12c isn't out yet, so when Fusion first came out, and even today, we implement multi-tenancy at the third level, what's called the infrastructure level. The infrastructure is the operating system. We implement multi-tenancy at the VM level. The VM level, the great thing about implementing multi-tenancy and being able to have lots of customers' data, efficiently using hardware and software resources. When you implement at the VM level, there's one great thing that happens.
You implement multi-tenancy at the platform level, the database level, all database security works properly. You implement multi-tenancy in the application, lots and lots of database security doesn't work at all. If you don't think security is important, that's fine. Back in 1998, 1999, that's fine. Implement multi-tenancy at the application level. You have no choice. That's pre-VMware. That's way pre-Oracle 12c. I thought then and I think now, multi-tenancy at the application level. Multi-tenancy at the database level is fabulous. You get highly efficient use of hardware and software resources. You can have multiple customers' data in the same database, but the database is responsible for separating it into databases in the same container database. All database security works properly. You get all of the efficiencies, everything. All the query and reporting tools work properly.
With Salesforce.com, even though they run an Oracle Database, because the Oracle Database security doesn't work properly, they have to implement their own security. Standard query and report writing tools don't work when you implement multi-tenancy at the application layer. Again, I forgive them. They did it in the late 90s. Workday did it recently, which is really crazy. VMs were around. Workday knew that VMs were around, but they did it anyway. It's very important to note that a big difference between us and, let's say, a Salesforce or a Workday. We're also a platform company, so if we think multi-tenancy is important, we can put it in the right place. We think VMs are important or networking is important. Technologies are important. We can implement it at the infrastructure level.
We're the only cloud company that delivers technology at all three layers of the cloud, not just in applications, but the underlying platform and the- If you're just an application company, you're going to have a problem because when companies buy your application, let's say they buy HCM from Workday, and they want to tie it into ERP from SAP. Go check out the Workday platform. Let me tell you about the Oracle platform. The Oracle Database, number one database in the world. Fusion Middleware, the number one middleware in the world. The Java programming language, the most popular program. Ask Workday about their platform. All they can do is do some simple file sharing. As we pursue the cloud, we're the only one offering intellectual property and innovation at all- Our competitors can't do that, and our competitors are too small.
Too much niche players ever participate at all levels of the cloud. You have more to your question, Rick? Or do you give up?
Hi. Can you talk a little bit about HANA and where you see it fitting in the market, and why you think they're selling as much as they have? Simple question.
Selling as much as they have. I love that question. Who are the HANA customers you've talked to?
A survey of 20 HANA customers and depth discussions, if you'd like it.
No, no. Go ahead. I'd love you to talk about the HANA customers. What do they do? Who are they? Don't need to tell me who they are. What are they doing with it?
Basically running analytics on it of SAP applications.
Do a lot of-
I'll send you the survey, though, seriously.
Okay. Let me say that we get a complete list from SAP of all their HANA customers. They have a contractual obligation to supply us with a complete list of all their HANA customers. We're still waiting. We can't find anyone in production with HANA. Maybe I'm mistaken. They claim to have 600 customers. Let me start with, it's not a database, the Oracle Database. It's much closer to Exalytics business intelligence, what you said, business intelligence. It is not a database. It doesn't do transaction processing. I think, again, you really shouldn't take my word for it because I think HANA is a project. We've been in the database business for a long time. They got a few guys in the lab that built this very quickly.
They must be like, you know, they're figure they're at least every one of them is 100 times smarter than our best guy to have databases. Exadata X3-8 has 26 terabytes of main memory, and we do database processing and memory. For years, we've had the TimesTen database, which is the leading in-memory technology on the planet. We've been working on this for a very long time. The idea, again, I think you just all need to go out and talk to these HANA customers and find out what they're doing because we can't find them. Not only that, we think what the for HANA would be they take a problem like MRP or the business ware to data warehouse, and they do an application, in other words. Not to try to build a general-purpose database.
They take a specific SAP application, like the business data warehouse, and they re-implement some of the query, and they make it work on HANA. HANA doesn't support SQL as we know it. You have to write custom applications for HANA. If you want HANA to work, you got to write a custom application. You want Exadata to work, you have to plug it in. Every existing Oracle application works unchanged on Exadata. Nothing works on HANA unless you write it for HANA. Let's start with that as one difference. Most of the databases on the planet Earth are Oracle. HANA. That would cost, I don't know, trillions of dollars. I know they got their own fund to help support HANA. They announced $150 million. Needs to be maybe $150 billion or $1.5 trillion applications specifically for HANA. It's crazy.
Maybe they can move one or two or three SAP applications to HANA. That's all they can do. You think the world's going to stop doing what they're doing and start writing stuff for HANA? Is this HANA? These guys aren't writing in the proprietary HANA language. All right. That's one. Two, to be a database, you have to deal with persistence, error recovery. Lose all your data. That's a problem. People get really upset. They get really upset, like, "Remember that data you had?" It's gone. They've been talking about this for a long time. I think it's great. I am so happy they're focused on HANA and not the cloud. This makes me very happy. They could have been rewriting SAP for the cloud, but instead, they built HANA. They haven't even begun rewriting their applications for the cloud. We started in 2004.
It took us eight years. They haven't started yet. In fact, in their last conference, they said they're going to start rewriting their applications for the cloud. Instead, they're really excited about this thing called HANA. HANA, to me, is much closer to what we call Exalytics, except for one difference. If you want your Business Intelligence applications to run faster and in memory, you plug in the Exalytics box. All the Oracle BI stuff, Hyperion times are 1,000 times faster because it's in DRAM. All the existing stuff runs. You don't have to rewrite anything.
What you're saying is, "Larry, HANA is so good, it replaces Oracle, and Exadata hardware, the Exalytics software, all the Oracle Business Intelligence, and the Exalytics hardware, and the InfiniBand network that connects it." If they can do that with 100 guys, they should run us out of business. I'm saying it's a joke. Conversation with you guys, you told me NetWeaver is going to wipe us out. You did. I mean, a room full of you guys saying, "Oh, you guys are so screwed. SAP's doing NetWeaver middleware. It's awesome. We talked to this guy." He had great hair. He said it was awesome. Really awesome stuff. I said, "You better talk." I mean, SAP going into the database business and the hardware business against us? Here we are. Make my day. It's crazy. It's absolutely crazy.
The favorite part is, what are they doing with their applications? I thought they were an application company. Tell me about R/3 or whatever their latest stuff is called in the cloud. They bought SuccessFactors and Ariba, and that makes them a cloud company? "Hey, we bought Taleo, and we bought RightNow." That's not what makes us a cloud company. That helps. What makes us a cloud company is we wrote all of Fusion for the cloud. We wrote the 12c database for the cloud. We've got all this fancy infrastructure for the cloud. We have our own cloud technology. We use OpenStack," blah, blah, blah. What have they got? This thing called HANA, which we can't find it working anyplace. Please, don't take my word for any of this.
All of you write a really great article about SAP, get their top five HANA customers. Go visit them. Find out what they're doing. See if you can find anyone in production, anyone doing anything useful with it. Yes.
Larry, you don't strike me as somebody that's easily satisfied. What can Oracle be doing better? What's your vision for the business five years out?
I'm a kind of mellow guy. I don't really worry about that stuff. I think the thing we're doing right now, the hardest thing that we're doing right now is actually now that we've made this huge investment in Fusion applications and the 12c database, now that we've built all of this technology for the cloud, geez, that's great. We made that engineering change years ago, right? Five years ago with the 12c project and the database. Eight years ago with the Fusion application project. Now we have to turn the field, that's our consulting organization, our sales organization, get them used to much more rapid implementations, selling applications in the cloud, which means not just saying our applications are better, but also explaining the importance of the platform, the importance of the underlying infrastructure.
Keep in mind, when you buy a SaaS application, what comes with it is an underlying platform. What comes with it is underlying infrastructure. All of that is important. It's not like the old days where you bought an SAP application, and you made a separate decision as to what database you're running it on. That's not the case in the cloud. In the cloud, you buy an application, a database comes with it. An operating system comes with it. Networking comes with it. Storage comes with it. Recovery system and security comes with it. All that's together. You make that one decision of this application, and everything else comes with it. It's very different. There's some things that are easier, that we, the provider, make the capital investment, you pay a monthly fee, all of that stuff. The implementations should be smoother and faster.
It should be smoother and faster because everyone runs the If you're at salesforce.com, let me tell you, what percentage of salesforce.com customers run on the Oracle Database, do you think? Simple. They do 100%. You don't have to worry about supporting DB server and all these things that traditional application companies did. A lot of the complexity and a lot of the variation is taken out of this. You should be able to do an implementation faster. You should be able to do testing more quickly. You should be able to come out with releases more quickly. It's kind of up-tempo, getting the consulting organization used to very up-tempo, rapid implementations, faster at a lower cost. Less variation, more standard processes implemented. It's a little bit different way of doing business, and we got a lot of people.
Turning the ship, you turn the helm over, and it takes a while before the bow starts moving on the horizon. We've got a lot of people to retrain, and we've been doing a bunch of that. We've had to reorganize the sales force. Against our secular competitors. We have an HCM Salesforce separate from our sales automation Salesforce, separate from our service automation Salesforce, separate from our ERP Salesforce. We have a lot of separate Salesforces. We made a bunch of changes. We line up against the competitor. We understand the product very well. We understand the competitive landscape very well. We've done all of that, and we really did that at the end of last fiscal year throughout this Q1 we've just finished.
Changing the company's orientation from on-premise software to stuff delivered in the cloud was more than just a bunch of engineering decisions we made between eight and five years ago. It's now the entire field organization has to be retrained and retuned and reoriented, and we're going through that right now. We're used to that. We started off building databases that ran on dumb terminals, then we ran on minicomputers and mainframes, and then along came personal computers and client-server architecture, Unix, then Linux, then the Internet. We've gone through these architectural changes, these changes of technology in the past, and we've been able to adapt first our technology and then our organization to these changes. I think we've done pretty well. We keep growing. I think, we've been through this before. The changes aren't easy, we've done it.
We're very focused, I think we're going to do it again.
How have the customer conversations changed in terms of moving to the cloud versus where we were a year ago, from your perspective?
The biggest thing, I think, was what caused us to introduce the private cloud. There's lots of about you're a big bank, and to take a lot of your data and put it in someone else's data center that you might know very little about. There might be regulatory problems. If that data center is located in the U.S., for example, there are different privacy laws in the U.S. than in the Nordic countries. Where the data center is located, where your data is located is important. What the disaster recovery plans are in case that data center goes down. All of that stuff is important. Our largest and most advanced customers who are used to asking a lot of detailed questions about security and disaster recovery were a bit uncomfortable just saying, "Hey, here's my data.
You do everything." They loved the idea of, if you will, the simplicity of the cloud, the fact that they pick an application, and they can be up and running pretty quickly, that they don't have to make a capital investment, that instead they got a simple monthly fee. It was good that they didn't have to make a lot of decisions about the underlying platform and infrastructure. They kind of liked being relieved of making those detailed technical decisions. On the other hand, it made them kind of nervous when they started asking questions about, "Well, Mr. Cloud Provider, what decisions have you made? What is your disaster recovery plan? What is your security model? Can I have a hot standby database? How fast can you get me back online?" Et cetera.
We think in a much more comfortable way to these large clients by saying, "Okay, we'll just have a piece of our cloud. We'll take the same exact infrastructure that's in our cloud, which is Exadata and Exalogic. We'll take those 2 machines, we'll install them on your floor. We'll own them. No capital investment. We'll run them. We'll manage them. It's all our labor running all that stuff. It'll be the same application, same database, same everything that we run in our public cloud. It will be on computers dedicated to you behind your firewall on your raised floor.
We manage it, we pay for it, and you pay for what you use." That has made our customers much more comfortable with the idea of taking advantage of the things they like most about the cloud without dealing with the uncertainties and what kind of disaster recovery and security they provide. I think that was the biggest thing that we found with our major customers, this general notion that, "My God, can I really turn all my accounting data and all my customer data over to some, especially when they commingle my data with my competitor's data in the same database, and there's no security barrier at all other than inside the application?" I'm going to rely not on the Oracle Database for security.
When you're using salesforce.com, I told you, security is relying on is salesforce.com's application layer security that salesforce.com built into their Salesforce automation application. That's what you're relying on when you use that stuff. A lot of people were uncomfortable with that. That's why we introduced the private cloud. We think that's a big deal. We think that is providing a private cloud. Having said that, let me now add, when they get the private cloud, it's very clear to them that they're making this decision on not just an application, but also the underlying infrastructure, the underlying database, the underlying middleware, the underlying business intelligence, all of these things. Again, you pick an application, all that stuff comes with.
I think they're much more aware of the fact that there's deep implication when you make one application, you want to hook it up to anything else, you're going to have to go into what salesforce.com calls Force.com. You're going to have to start programming in Force.com to connect that Salesforce application up to whatever else you want to connect. You're programming on a proprietary platform called Force.com. You're proprietary. You're completely reliant on whatever security model salesforce.com implemented in their application. You're totally reliant on whatever disaster recovery salesforce.com has. We think to our prospective customers for cloud technology by describing what we offer as infrastructure, what we offer as a platform, and what we offer as applications. Now I should go. No one? It's a very quiet section over here. Okay. You don't have to ask a question.
I just asked a couple earlier. I have a hard question about your hardware business. Do you really have two Exadatas in the boat that follows your catamaran?
No.
I'm kidding.
Maybe we do, and I don't know about it. I kind of doubt it.
More seriously, if you look
Oh, that was a joke. Okay. I didn't realize the joke. I was wondering, I actually just thought to myself, is it possible we have two up there?
I mean, I have heard that, I don't
Yeah. We have a lot of telemetry signals. We're taking lots and lots of data across a radio network from the sailboat into our telemetry boat. I don't think we process some of it real time, but a lot of it is saved and processed as it is. It's unimaginable to me that we have two racks of Exadatas.
More seriously, if you look at the hardware business, we talked a lot about Engineered Systems today. They seem to be doing extraordinarily well. If you look at the portfolio of your current hardware products and try and think two, three years out as you always do, what do you see? What's your over the next two to three years? Where do you want to invest more, where do you potentially want to pull back and invest less?
Businesses. We have one hardware business where our intellectual property, where we add lots and lots of value, Exadata, Exalogic, SPARC SuperCluster, our T4 Unix computers. We add lots and lots of value. Our InfiniBand switches, these kinds of things. Our ZFS Storage Appliance, we have a huge amount of intellectual property. We make huge investments in this. That's one of our hardware businesses. We have this other hardware business where we take an Intel microprocessor, Samsung memory, an operating system from Microsoft, we sell that business. We have these two businesses. We have the IP-rich business, and we have the commodity hardware business. The commodity hardware business, when we took over Sun, it was quite large. We had disks from LSI. We got out of that business. We had disks we bought from Hitachi. We were a reseller. We had no value.
We got out of that business. We are not a distributor. We are not like HP. We are not like Dell, not even like parts of IBM. I mean, it's okay. I mean, if Dell does a really good job in the commodity business, I mean, there's nothing wrong inherently with being in that business as long as your cost structures and your distribution model, indirect model, and you do all the things right to be an effective business you want to be in, and you know how to do it, you know how to control costs, and you know the margins are going to be thin, and you can make money. I have no problem with commodity business as a concept, just not our business. We have been exiting the commodity portion of the Sun hardware business.
We came out with a T4 microprocessor, which is very, very fast. I kind of made the promise with our partner, Fujitsu, we're going to be delivering the fastest microprocessor in the world in the not-too-distant future. We think we'll run much faster because our projects are to move things out of our software directly into silicon on the chip. That's interesting because we're in a good position to do that. We're a software company. We're a silicon design company. We can do that. We get a huge like that business. We're an IP company. We develop IP, and we sell differentiated products based on that intellectual property. That is our business. As you look at these two hardware businesses, the hardware business, the ball constantly going down. What is the problem?
Well, these commodity businesses are shrinking faster than the businesses we care about are growing. Exadata, Exalogic, Exalytics has been growing in excess of 100% for a long time. T4, the SPARC T4 Unix machine is the fastest growing Unix machine in the world. ZFS Storage Appliance, a competitor to NetApp, is growing extremely rapidly. All our IP. Our IP-rich hardware business is about to cross in size our commodity hardware business. At that point, we started out with a small IP-rich hardware business and a large commodity business. Small numbers doubling. Eventually, that small number keeps doubling, and it's bigger than your commodity business. That happens this fiscal year.
This fiscal year, our Exadata business will be vastly larger than our x86 business, or our x86 and commodity storage business combined, and the IP-rich business and the overall hardware business will start growing pretty rapidly starting this fiscal year. Our strategy is that we're an IP company. More silicon, not just Apple. Again, the closest company in the world to Oracle in terms of philosophy or the closest company I've admired the longest is run by my best friend of 25 years, Steve Jobs. He believed- own microprocessors. They design their own displays, and they design, obviously, their own operating systems, blah, blah, all the way up. We believe we should be doing the same. We shouldn't be in the business with our intellectual property. I love the hardware business if it's differentiated. Now, one last thing about Exadata, Exalogic, Exalytics, SPARC SuperCluster.
I think the general-- our tone has always been, God, these things run unbelievably fast, 10, 20, 50 times faster than- is the people that need extreme performance. That's a decent market, but not the most important market for Exadata, Exalogic, Exalytics. The most important market for those products are people that want low cost. Exadata box will run more databases at a lower cost than a room full of commodity x86s or Xeons, whatever you want to call them. I believe that's true. We're not just going after the high-end market, including rooms full of x86s. We think the easiest part of that market to get are the extreme performance part, the people who just need extreme performance. We think those are the early Exadata customers.
We think the systems at the high end of the market, the bulk of the business as we grow two, three, four, five years out, is going to be replacing rooms full of x86s, which consume enormous amounts of power, are very difficult to manage, are very labor intense. A small number of super servers than huge numbers of small servers, as long as the cost-performance equation works. We think it does. We think Exadata is much cheaper to run than a room full of big. The server business is $50 billion, and Exadata is half a server and half storage. The storage business is now slightly larger than the server business. They're all around 50. I think that's the $100 billion market that Exadata is in. Of the extreme performance portion of that $100 billion is probably only, I don't know, 15% of it.
Hi, Larry. You've been talking about how you're making this transition from selling the components to selling really complete solutions over the cloud. You've got a large portfolio of applications that you've been supporting. A lot of them have been acquired, and you look forward. I'm interested to hear your thinking about how long will it take before, say, the majority of your customers who are running applications are ready to make that move to the cloud, in that process, what do you support? How long do these long tails either push or pull the demand to try to make that transition?
One thing is interesting. We love our existing customers, and we're not going to try to coerce them or push them to move to Fusion Applications. Let me make a very important distinction here. I think this is really important. Our traditional applications like PeopleSoft or E-Business Suite on a private cloud or on our public cloud, on infrastructure. When you say move to the cloud, you got to be really careful about SaaS applications, cloud SaaS applications equal. Moving to the cloud would mean, for Oracle, selling a Fusion Application. I think that was how you're going to couch your question. I submit to you that if you take your existing E-Business Suite application at the infrastructure level, and you're now running a traditional application, just like you moved to amazon.com. You can run that at amazon.com.
You lift it right up right now, move it to amazon.com, you're running in the cloud. You can take all of our E-Business Suite, et cetera, can move those applications onto our infrastructure as a service, either in the Oracle Public Cloud or a private cloud behind your firewall. They don't necessarily have to move to Fusion to get them onto the cloud. We think that we have a lot of opportunities. That's one of the big uses of infrastructure. It's one of the big. Compete with Amazon. Well, yes and no. Yeah. We're in the infrastructure-as-a-service business, but we're using Exadata, Exalogic, Exalytics, which we think is faster and cheaper, but other people can decide. We think someone who's running, say, E-Business in the Oracle Cloud than the Amazon Cloud.
The kind of services we provide, the kind of backup recovery, disaster recovery, backup services, et cetera. The fact it's one vendor, it's us, makes it easier for them. We think we have a built-in advantage as we compete with Amazon or Rackspace or some of the other infrastructure-as-a-service companies. If you look at our opportunities in the cloud, it's not just selling Fusion Applications Tal eo right now. It's also selling this infrastructure as a service to our traditional customers, and they can take their time. We can get a lot of the benefits. An application change. We get benefits as a provider, as a service supplier. They get benefits as a consumer. It runs faster. They don't have to upgrade their hardware again, their operating system. We can take over the management of the application for them, upgrade it, do all of those.
We think our ability to grow our cloud business is going to be measured not just how much Fusion we're selling, but how much infrastructure as a service we're selling to our very large install base. That's why we're really optimistic about how fast our cloud business is going to grow. Right in the middle.
Just maybe coming out of the clouds, coming down to the products and talking about the database. It's been, I think, 5 years since you've had a major new database launch, I think with nine- As a license. I'm just wondering if you could maybe at a high level, talk to us about 12c and how you look at that as an opportunity for that core database business versus the past few major releases you've had over the last decade or so.
Yeah. Well, again, we made the decision because we decided to make 12c a container database with pluggable databases, in other words, make it multi-tenant. It took a long time to do. It took a lot of re-architecting, and obviously, we had to make security work. They had to make everything work transparently, right? As you go from Oracle to 11 to Oracle 12. For all the SaaS companies. We think there are going to be lots and lots of SaaS companies. When I say a database for the cloud is not simply rebranding, it doesn't take us 5 years to change an 11g to 12c if it was just branding. That would not take 5 years. We did a lot of re-architecting, and we think it is the right platform to this whole next generation of SaaS providers, and not just us.
It also has tremendous utility inside of our large customers who want to consolidate lots and lots of separate databases. They've got a room. The whole idea is everyone's gone to this model of rooms full of x86s, rooms full of Xeon servers running Oracle. There are just lots and lots of these. It's at the point, I'm not sure how aware everyone is of this, the electricity to power a Xeon server over 2.5 years costs more than the Xeon server. The power costs more. This notion that they're cheap, and then the labor associated with each one, backing up each one, managing each one, it's enormously expensive. This whole idea of manage many as one You can take 1,000 of these Xeon servers and consolidate them into four Exadatas or 10 Exadatas, whatever, 10 racks.
It's not just the power savings and the floor space savings, but the management. You have to back up 10 databases, not 1,000. You can restore one at a time. You can recover one at a time. You back them up, you manage them as 10 databases rather than 1,000. Has huge implications in terms of labor savings, reliability, floor space savings, electricity savings, et cetera. The world used to use a small number of large machines. Now it uses a large number of small machines, and I think we might now oscillate back to a small number of large machines because it's just going to save you more money. More money on labor, more money on power, et cetera. That's what we think is going to happen, and we think that's going to be a boon to our database business.
That was the decision we made. Those are the two driving forces behind 12c, those two business opportunities. Way in the back.
Thank you. Kash Rangan, with Merrill Lynch. Larry, you asked us to find a HANA reference. I found it. It's in the island of Maui.
Maui?
Yes.
Good God. It's only eight miles away from Lanai. Where exactly on Maui?
The question is it easier to drive on the road to HANA or sail to HANA? It's a serious question.
I've driven on the road to Hana. Hana means flower, by the way. I don't think that's what SAP means. I think for SAP, it's an acronym, but I think the Hana in Maui is flower.
I had a serious question, actually, if that was not a question.
I consider that very serious.
Okay.
Go for it. Go ahead, Kash.
All right. The question is, late '90s, you talked about the common database, ERP, CRM, supply chain, all running on one database, and I thought that was a pretty powerful message. Looks like I was talking to Thomas, and he said that Fusion apps also run on one common database. Your competition, the larger applications companies, really install multiple modules on multiple databases. It looks like you have a powerful cost of ownership argument in the next generation of applications. Can you talk more about that? Thank you.
Yeah. We think, again, this whole notion of having the more of your company's data, let's say, data integration, we'll call it data integration, is a good thing. We think if your information about all of your customers is in one database, you're a big bank or a big insurance company, having all of your data about all your customers in one big database is good. Having it in 50 separate databases, fragmenting it, is not good. There's just this general notion of you want to do cross-selling. You want to sell your insurance customers banking products. You want to sell your banking customers insurance products. If you have totally separate systems for insurance and banking, it's harder to do that.
If you have sales automation products and accounting products, and you want your salespeople to cooperate with your legal people in drafting contracts, and they're in separate databases on separate systems, it's hard to build that application. The general notion, the fewer databases you have, the more consolidated your information is, the better off you are. You'll see all of our systems, even though you can take an Oracle HCM system and hook it up to other things. If you were designed, if you pick the whole Oracle suite, all of your data is integrated. In other words, it's not fragmented. All of your data is in one place. All your customer data is in one place. All your employee data is in one database. All of your product data is in one database. Therefore, it's very easy to ask a question and get an answer.
It's very easy for one group of people in your company to work with another group of people in your company because we all share the same system. We're all in the same database. It's much lower cost to manage one database to back up, one database for disaster recovery. Again, it doesn't have to be one, but I'm saying one is better than 10 is better than 100 is better than 1,000. Avoiding data fragmentation and system fragmentation is a very big deal. That's one of the reasons why we're in the suite business of applications. A lot of people envision, "Gee, I'll get 30 different applications from 30 different SaaS providers." It can be very tricky to hook up all those pieces. A lot of your information is going to be fragmented on 30 different databases.
It's going to be very hard to find out what's going on inside of your business. Suites, the idea of having a lot of things that are engineered to fit together. We believe there's horizontal suites, like at application layer. CRM should fit with ERP, should fit with HCM. There are vertical suites, if you will. The application should fit with the middleware, should fit with the database, should fit with the operating system and the VM, so I can patch all of it at the same time. One of the really cool things about the Oracle Cloud, or sometimes people call it Red Stack, if we send out a bunch of bug fixes or an upgrade, you don't separately upgrade them. Think about what goes on when you have 30 different applications on 30 different databases, and how you fix bugs in all these different places.
We send out one file that patches all the applications, database, middleware, Java programming language, the VM, the firmware, the operating system, everything in one file. You know what? We've tested it all together. We've tested it. You don't have to test it. We've tested it all. It's much better if we test it all together than if you test it all together. This whole idea of fragmentation, taking a lot of little pieces that were never designed to fit together, and you buy them from 30 different vendors, you're responsible for doing the integration. You're responsible for doing the patching and the bug fixing. You're responsible for fitting all these pieces together. Then when fixes and upgrades are sent out, you have to fit them together again, and again, and again.
That's why being a big company is an advantage in this case, because we can take on the entire problem, a complete suite of applications, a complete stack of technology. No data fragmentation, much lower cost of ownership. All your data's in one place. Back up one database, one patch file for everything. We think it's a huge competitive advantage and something that would be very difficult to do if you're a small company. Over here.
Hi, thanks. Phil Winslow, Credit Suisse. Larry, we've seen a lot of companies start talking about sort of a revival of some old data management systems and also some new ones, object-oriented, NoSQL, et cetera. How do you view these as whether they're threats or opportunities for Oracle? For example, do object-oriented databases actually scale up to high-volume transactions? What's the role of NoSQL? Just how do you see this market evolving?
Well, we have a NoSQL database. We built object technology into the Oracle Database a very long time ago. Oracle handles unstructured textual data and searches it. With Endeca, we bought a bunch of algorithms that's going to allow us to continue to improve our management of unstructured data. Oracle's not a relational database, exactly. I mean, it's a relational database, too. It started as a relational database, but we handle structured and unstructured information. We handle objects. We handle images. We handle all sorts of data. We think it's very important to have all different types of data inside the Oracle Database. Now, if you have a very specific need, like in the old days, an in-memory database, TimesTen was highly specialized in that niche, extremely important. We think it's great technology, we use it.
We have a NoSQL database, which we use in our Oracle Big Data Appliance. It's not that these other products aren't important, but they're niche products, and they're really no threat, if you will, to the Oracle Database. I would say HANA. Well, I say it's a joke, but it's actually, to be fair, let's say they do a good job of it. It's a niche product. It's a niche product that's going to be useful for doing data analysis for applications written specifically for HANA. It's a niche product. It's no threat to Oracle whatsoever. In a lucid moment, I'm sure Hasso would admit they'll never be able to do high-performance transaction processing. Never. It will be in that interesting niche. Maybe they'll do a good job of moving their own MRP on top of HANA and do high-speed in-memory MRP.
We look at that as a competitor, not to the Oracle Exadata machine, but we look at HANA as a competitor to Exalytics. Not even a general-purpose competitor to Exalytics, a niche, an Exalytics niche because Exalytics handles all sorts of general-purpose business intelligence. HANA, you have to custom code to. That's not true of Exalytics. Exalytics speeds up all of your existing stuff. They're somewhat niche products. Oracle Database has gained share every year for, I don't know, the last 25 years, 30 years. I mean, it's for a very long time. It's called the information age. It's not even called the cloud age. Before that, the PC age or the tablet age or the smartphone age. It's not called any of that. It's called the information age.
We have the number one piece of technology for managing the bulk of the world's high-value information. Over here. Then in the back.
Thanks. Kirk Materne with Evercore. You mentioned that scale-out architectures, you think they're going back towards Engineered Systems over time. When Facebook, Google were building up, Engineered Systems weren't around. I'm just curious as to what you think sort of inspires them to start thinking harder about Engineered Systems. When you guys talk to large public telcos that are going to build out their own clouds-
Yep.
How do you all make sure that the next set of public clouds include Engineered Systems? Is it an either/or thing? Can you have the scale-out architecture plus the Engineered Systems because-
The Engineered Systems are purely a scale-out architecture.
Okay.
That's all they are.
All right.
Right? I mean, the Engineered Systems, the reason we use InfiniBand, that's what it is. It's a bunch of x86 machines loaded up with flash and DRAM connected with InfiniBand, and there's more to it than that, but it is a classic scale-out architecture. It's not a scale-up architecture. All of the Exa stuff, SPARC SuperClusters, everything is scale-out. That's why there's no single point of failure. It's why you add capacity, you just plug in more nodes. You plug in more and more nodes into Exadata or a supercluster. That's all scale-out. But the fact that we use a faster networking, that we use very different software, in some cases very different silicon, allows us to run this scale-out network faster than anybody else. Therefore, we should use less hardware.
The argument to a Facebook or a Twitter to move away from commodity boxes to Engineered Systems is twofold. You'll run faster, you'll spend less. That's pretty persuasive. I spend a lot of friends over at Facebook, including Mark Zuckerberg, We talk about this all the time, and again next Monday. We'll see. We want to win two parts of that argument. You'll run faster, so you can offer your customers all sorts of insight from all this enormous database you have. Now that you've passed the 1 billion regular user count. Just same as Twitter. You have a chance to monetize this enormously valuable database. If you use our Engineered Systems, you can provide a level of insight you can't provide with a bunch of commodity boxes. Oh, by the way, you get that insight only if you're willing to spend less.
That's our argument. That's our argument to Mark Zuckerberg over at Facebook and the guys at Twitter, Also every large organization with lots and lots of databases all over the place. Lady right there. Yeah.
Hi, thank you. I was just wondering if we could get your comments. In the years pre-Y2K, there was this massive uptake of packaged apps, SAP got a big lead in the market. With these apps now facing the kind of 10- to 15-year useful life, I am just wondering if you see a lot of market share up for grabs over the next 3 to 5 years in the app space. If so, what type of opportunity does this create for Oracle?
Well, there's a famous saying, "Those with the most, have the most to lose." I'll just ask you're the CEO of SAP, you think cloud's important? Maybe. Well, why didn't you move a single application to the cloud? Isn't that your primary business? Instead, they built this thing called Business ByDesign, a colossal failure. It was an idiotic idea because it was designed to sell basically ERP to companies with less than 100 employees. Now, does the existing SAP sales force call on companies with less than 100 employees? No. Does SAP have a lot of brand equity with companies less than 100 employees? No. Can we reuse any of our existing technology to pursue this new market, companies less than 100 employees? No. We can't use the sales force, we can't use the technology, we can't use our brand.
Why the hell are we going into this business? It's very strange. Business ByDesign, a colossal failure. They need to rewrite all of their existing applications for the cloud. They haven't done it. They bought Ariba and SuccessFactors, and they said they're going to have nothing till 2020. I like our chances. Now they're going to melt away like a glacier. Everyone predicted the end of IBM, the end of mainframes, right? Well, mainframes are still around, and they're less important every year. I don't think there's going to be this evacuation away from SAP R/3 and all this other stuff overnight. They have nothing to offer their customers until 2020. Nothing. It's a long time in the computer business. As Fusion Applications mature, we think we're going to be able to pick off a lot of their installed base.
We're very optimistic that that's going to be one of the companies that really is having a very difficult time. I'll ask you, if you're SAP, what do you do? Have they even said they've started rewriting all of their applications for the cloud? Have they made that statement yet? What are they going to do, just ignore it? I'm asking you, anyone want to explain it to me? You're the CEO of SAP. You haven't started yet. What are you going to do? I'd love to hear one idea of what you do. Yeah? Okay. Now, the problem with that is predicting exactly when they fall off a cliff. Why would a new customer ever consider how they sell into an install base? They got a huge install base. They're the number one application company in the world.
They got a great brand in the oil and gas industry. They're a formidable company. They talk about, I know that you say, "I don't get the cloud." I don't get the cloud. Does McDermott get the cloud? Does Snabe get the cloud? Does Hasso get the cloud? If so, what are they doing? What are they doing? Right here.
Larry, I want to switch maybe from a technology and competitive question to something.
I'm so disappointed.
Thought you would be. If I roll the clock back, a lot of years, maybe around 1990, I remember Oracle launched a big drive to become more customer friendly. I remember product quality was a big part of that, I think there were other parts of that too, as I recall. Since that time, Oracle's obviously changed a lot. I'm sure the development processes have become much more rigorous. You don't hear about quality issues. I'm wondering, do you see a risk as the company grows that it gets harder for customers to do business with Oracle simply because it's got more products, it's got a lot more customers? Is there a risk Oracle becomes more difficult for customers to deal with? If so, what do you do about that?
This would be the greatest question of all time. Are we too big? I don't think so. One of the things we try to do is we broke up the sales force into separate sales forces by specific product area. I think there is a danger if you get to be so big and you have so many generalists. You can't have the same salesperson selling InfiniBand switches and budgeting and planning for the CFO and InfiniBand switches to the networking manager at a large company. Just doesn't work. I think the biggest danger is we have so many products, and we have people selling the products that don't know much about them and don't know much about the competition. We've decided that lots and lots of separate, highly specialized sales forces associated with highly specialized sales consulting and consulting teams is the way to handle it.
You kind of atomize the company. You take this big company, and you break it up into areas of interest. There's an application group, the application group itself is divided into CRM, HCM, and ERP. ERP itself is divided up into supply chain, manufacturing, and so on. You keep specializing. You keep atomizing. You constantly measure yourself against your secular competitors. Our sales guys are measured by how well they do against salesforce.com, for example. Our hardware guys are measured how well they do against IBM. Our infrastructure guys, we measure how well they do against Amazon or Rackspace. You stay very focused on what is the technology? What is the market? Who's the competition? Who are the potential customers and the existing customers? How do we gain competitive advantage? How do we gain share? You measure everything constantly.
In terms of doing business, I think this whole move to cloud actually makes it easier to do business with us because your customers said they didn't want to make a capital investment? Don't have to make a capital investment. You got plenty of capital. Mark Hurd was telling me, one guy said, "I don't want this cloud. I got tons of capital." Well, you don't have to buy the private cloud. We'll still sell you an Exadata, but you get a choice. We have more modes of doing business, which I think makes us, as we give customers more choices, I think just the opposite happens. We become friendlier or more accommodating to doing business the way they want to do business.
We're talking to a very large customer right now about not only a private cloud, but us running the infrastructure as a service, but also us being the DBA for all the Oracle databases and kind of a little kind of combination private cloud writ large, a little bit of outsourcing to go with it, at least for database services. We'll do that, too. I think we're learning to be more flexible in terms of giving the customers more choices and giving the customers suites of products and services the way they want to buy them. I hope just the opposite is happening. We're staying specialized. We're lined up against our competition. We understand the technology, and we have lots and lots of different ways of delivering the technology to our customers. Thank you very much.
Well, thank you all very much for attending today. We very much appreciate you taking the time out. Just as a reminder, you have surveys. If you would please complete those. You can just leave them right where they are. That'd be fantastic. Thank you all for coming. Have a great night. See you next year.