Good morning, everyone. Thank you for joining us here for Seagate's Strategic Update. My name is Kate Scolnick. I'm the Vice President of Investor Relations. Just a few things we'd like to cover here before we get started. First and foremost, our forward-looking statements. We will be making forward-looking statements today, and we refer you to the risks and uncertainties we have filed with the SEC on our Form 10-K, most recently filed in August. Right now, I'd like to bring on the executive management team to join me on stage. Steve Luczo, Patrick O'Malley, Dave Mosley, Bob Whitmore, Rocky Pimentel. I know, right. All front and center.
Based on the feedback that we have had from a lot of folks over the last several months, we formulated an update today, we really purposely called it a strategic update because we've had a lot of questions about the next steps for Seagate Technology, both from a technology framework as well as a financial framework. We've designed our discussion today with two main speakers covering a vast majority of the information, but also having the executive team here to be available for questions and answers. Your job was to come prepared with questions for them, and we hope to make this very interactive. The session is also being webcast, the session will be available via archive shortly after the presentation. We will start off the meeting with a strategy update from Steve Luczo, then open up to questions, take a quick break.
We'll be back at 10:30 A.M., an overview from Patrick O'Malley on Seagate's financials, then some more time for Q&A with the entire team. Before we get started with the entire intention of what we're here to do today, we do want to make a few comments about the current demand environment. We know everyone is very interested in that. As others have made comments about in our industry, we are seeing a more muted demand and some rebalancing of inventory that is causing our expectations for the addressable market opportunity to be more muted than originally expected. We do think that's going to be approximately 10% lower than we originally thought in the July timeframe. The impact to that to our top line will be less than that, probably approximately 5%-7% lower than we originally expected, and margins around approximately 30%.
We'd like to take a few questions on this information right now, then head into the full intention of the presentation. With that, I'd like to turn this stage over to Steve Luczo, our President, Chairman, and CEO.
Okay. Sorry, I didn't know all these people were going to be up here. I thought I was going to have five empty seats of fictitious analysts that I could talk to. You want me to do what? Yeah. Look, the purpose of this meeting isn't really to drill into this quarter, but just in light of the announcements by Intel, HP, Dell last week, WD, it maybe makes sense just to answer a few questions to provide a little more color about what we're seeing right now. Maybe just a few minutes on that, if there's any question. Rob, yeah.
How does it feel, quarter on market share? Is anyone using pricing to try to gain the market share back?
Okay. Can everybody hear the questions in the back? Christian, did you hear it? Okay. The question was, given the weakness in demand, what's going on with market share? Are people using pricing to basically gain market share in light of the weaker demand? Yeah, I'd say no. I think the market share figures are going to probably come in pretty consistent to what we saw last quarter. I think it leads to a bigger question, which is, if demand is not what the industry thinks going into a quarter, what happens now versus maybe what happened pre-consolidation? I'm going to talk to consolidation here shortly. Look, I think one of the biggest benefits of consolidation is really alignment of supply and demand.
If you look at the consolidation that's happened in the supply chain as well as in the industry, it just means that there's less chance of error that either parts and/or products get built ahead of recognizing what's going on with demand. It's harder to slow the system down. You have excess inventory that you have to figure out how you're going to get rid of. I think with the consolidation and also with the efficiencies of supply chain that certainly have evolved over time and even upstream with our suppliers and then through the customer base, there's just a much quicker response time. Fewer players, less component people overproducing, less drive companies overproducing, therefore, it comes into alignment quicker, so you don't get basically stuck with as much excess inventory that you might have to aggressively price.
I think the pricing mechanism is a much smaller factor than it would have been, say, in a similar case three, four years ago. You still have absorption issues. I think I'll let, when we get into Q&A, Dave talk a little bit about what's it like running a factory when you think you're making X number of drives and then halfway through the quarter it turns into some percentage of that. It's still challenging, but I think the impact in terms of gross margin impact is much less than it was historically. I don't see anyone in the industry trying to buy share to make up for the fact that demand appears to be down for some period of time. Far I'd say the industry's behaved pretty well there. Rich?
Good morning. Does the 140 TAM then reflect your reaction to the demand? Is that correct of rebalancing the inventory for the next quarter, or do you think that we need to have a little bit more of a pullback even, or a flat quarter sequentially in December to really solve the inventory issue? Is it a one-quarter phenomenon?
I think it's hard to say, Rich. I think down 10% is probably bigger decrease than what end user demand is, which means, yeah, I think there's an inventory correction going on, not necessarily maybe at the drive level, or not just at the drive level, but probably at the drive level and the systems level. I think that exiting the June quarter, the computer companies probably were carrying more inventory than they typically would out of June, and then it was into an environment that was worse than they thought. I think there was a lot of systems inventory that had to be worked through, as well as maybe excess drive inventory, depending on who the customer was, actually, that was probably not an issue at all on one end of the scale to being a bigger issue on the other end of the scale.
I think it is a rebalancing. I do think all things being equal, you're going to see an uptick in the December quarter. I think if 140 is the answer, whether that's 142 or 143 or what, I think you're going to see a TAM in December that's probably more in the 150 to 155 range. I think most of the inventory issues are probably resolved at the customer level, but we only have certain visibility on that. Rick.
Steve, when you look at the trends that are going on in the marketplace, whether it's tablets or Ultrabooks or what have you, what makes you so confident that the demand that you're seeing is cyclical in that it was macro in nature versus secular as you look for this quarter and beyond?
Well, yeah, I don't know that I said I was confident about that. I don't know. It's the right question. I think what all of us are thinking a lot about and planning around is, what is this all about? Is it macro? Is it something to do with a change in computing lifestyle? Is it something in between? My assessment is probably as follows. What my confidence is, I don't know. I think most of it is macro. If I think about, say, this quarter or the December quarter, I think when we were looking at this back in the early part of calendar 2012, we were thinking that the December quarter was going to be about 185 million unit TAM. Now, Seagate was on the low end, by the way. There was a lot of people that were in the 200 to 205 range.
If you're at 155 or you're 30 million units off, I look at that. If I had to splice it up, I think most of it is macro, maybe. Well, what does that mean? I don't know, 15 or 20 million units. I think at the margin is there cannibalization, pure cannibalization of tablets versus notebooks. I'm not as convinced. There could be some, at the margin, it could be maybe it's 5 million units even, which would be actually a fair amount. The question you have to ask yourself is, did someone buy a tablet and they threw their notebook away or they're never using their notebook again? I just don't know that happens.
It seems to me that tablets is another form of accessing data that has certain advantages versus a notebook and certain disadvantages. It's another device in the ecosystem, which from a storage perspective, by the way, is a really good thing because it's generating a lot of demand for petabytes. We like to think of all these things of, because people love the razor blade analogy. To me, notebooks and tablets and smartphones, those are just razors, and we're the razor blade. The more devices that are out there that connect people and have them share data and have to deal with bandwidth issues through caching means lots of disk drives. Interestingly, our retail business actually feels pretty good this quarter, which would be a little counterintuitive to the general theme of consumers being down.
It may be driven in part by because obviously tablets have no real memory on them, so if you want to do something with them, you probably need a storage device near them. I think that cannibalization is probably not that great. There's another type of cannibalization that might be greater, so another 10 million units, let's say, of the 30, which is it possible that in a weak macroeconomic environment that because I bought a tablet, I didn't buy a new notebook? I delayed a purchase of a notebook. I have a notebook. It's two years old, it's three years old, it's four years old. I normally would have replaced that notebook by now, and I didn't because I bought a tablet. I think that is going on now.
I don't call that cannibalization yet because I think it actually probably gets replaced once either the notebook offerings become more compelling or the economy gets better. On the first point, it's a fair criticism that if you look at the notebook offerings from the computer companies over the last three or four years, there's not a lot of compelling reasons to buy a new notebook unless you're buying one from Apple. I think that's about to change. I do think the thin and light, I do think, I'm not saying this as a Microsoft director, I do think Windows 8, I think touch and keyboards is definitely where the world is going. It actually kind of surprises me that OS X Lion didn't do that. I think that this is the front end of what's probably going to be a multi-year trend of more compelling notebooks coming out.
I think that's probably a big part of it. I think the majority is macro. I think whether or not it's Intel or FedEx, there's a lot of data out there that's just saying that globally, there's a pretty substantial slowdown going down, which obviously governments are trying to address by pushing rope. Who knows how it gets resolved and when. To me, the most telling one was Intel. For those of you who've been around for a while, you know that Intel is the last company that's going to see an order cut for a lot of reasons. They have long lead time requests. They have penalties when you cut. You can't get back in the queue if you want to, without paying a lot of money. It's not hard inventory to hold. It's chips. Take them and throw them in a drawer.
They're not going to get outdated. If I don't use them this quarter, I'll use them next quarter. If they're getting cuts, that to me, was really significant. For the drive industry, though, we counter that against petabyte growth that's very strong. If you've been listening to us, we view ourselves not as shipping units, but as shipping petabytes. The form that those petabytes take, in a certain sense, is irrelevant to us. The majority of our R&D and capital is in heads and disk, not in drives. Our easiest to move or change or flex is our drive capital and factories. If units are flattening, but heads and disk are growing, which they are because of petabyte growth, that's really good for Seagate because we're absorbing heads and disk, and that's really what we need to do.
The management team, I think, has to stay close to, is it $150 million? Is it $160? Is it $140? Is it long-term? Is it just for the next six months? It doesn't fundamentally change the bigger issue, which is petabyte growth is 2X areal density growth. Which means without more capital going into our industry, there's going to be some sort of supply-demand imbalance the other way, because we can't make enough petabytes on the path that we're on right now if you do an intersection at around 2016 or 2017. For us, it's how we package it. Yeah, we need to be smart about it. You don't want to have $100 million notebook drives ready if there's only demand for 72 of them. We're pretty smart about that. We see the shift to the cloud.
We see the shift to higher capacity per drive. People now will say, "If you could give me seven or eight or 10 terabytes under a spindle, I know how to manage it." Well, for us, that means that there could be six or seven disk under a spindle, with 12 or 14 heads. That's a really good thing for Seagate. We would rather make three times as many of those versus six times as many single-headed or one just two-headed drives. That's just where I'm thinking. We have to watch it. We'll see. I still think it's more macro right now. I'll take one more question. Yep.
Yeah, thanks. Given the demand environment and the TAM, we've heard you talk about 185. Can you also talk a little bit about how this has had any impact on what's been the long-term purchase agreements that you've been able to sign? Has there been any flexibility in that given the demand environment? If there's any update on those, that'd be helpful.
The question was, did the change in demand impact the LTAs anyway? No. The design of the LTAs was more about in a highly constrained environment. We wanted to align with customers who were more interested in a strategic relationship as opposed to a tactical relationship. If Seagate wanted to go full tactical, we would've just auctioned drives for nine months, and frankly, probably made a lot more money than we did. We felt that because we had a lot of customers who actually really were strategic with us even before the flood, and then a whole bunch that became strategic with us the day after the flood, we decided that we would turn that into an agreement that said, how do we work with each other over a longer period of time to manage our businesses to focus more on velocity than on saving $0.50 a drive?
Look, there's a set of customers out there that believe that supply chain means buying a part as cheap as possible. There's a set of customers that believe supply chain is, how do I get as much velocity through the system as possible to generate revenue? What's the difference? The difference is, I have a procurement organization that's really crafty, that makes Rob and I compete against each other, and I buy a disk drive $0.50 cheaper. On the other hand, Drew is running a company that's saying, "No, what I want to do is take a disk drive and turn it into revenue in three days instead of 12 days." That's $800 to $5,000 of revenue versus saving $0.50 on a disk drive.
What happened with the LTAs is it made people focus on where were their priorities, and they realized that the ones that were focused on velocity were going to take a lot of share. The agreements were not just about providing supply, it was what's the underlying relationship that's going to occur through the supply chain to their end customer demand. Some companies are better at demand shaping than others. Some have a lot more SKUs than other, which makes it very hard to change if my orders change. You all think these things are straight commodities, but they really aren't. Depending on which notebook I'm putting the drive in, it has to be a certain drive with a certain code that's been qualified. If you have more flexibility around that, obviously, you can generate a lot more revenue. The LTAs were more about that.
I think the Street viewed it more as this kind of thing about guaranteed revenues or guaranteed prices or whatever, which was part of it. That hasn't changed. We had flexibility around share or units. We obviously work with our customers. We're not going to make them buy inventory they don't need. I think consistently, we've over-shipped to the LTA volumes. My guess is we'll do the same this quarter. I don't really see any change through the end of the year as a result of the softening demand. We've had a couple of customers that have actually extended the LTAs, I expect that there'll probably be a couple more. These things are mutually beneficial to customers that think of us strategically. My guess is they stay in existence in various forms for a number of years now. Okay.
Let's move on to this presentation, then it gives you a little more time if you wanted a couple more questions when we get to Q&A. Again, I'd rather not have this all be about what's going on right this half a second, because the story is a little bit bigger than that, we believe. What I'd like to do is first just give a quick snapshot for fiscal 2012. Oh, I have to make a pitch for my Hungarian friends at Prezi. This presentation is done off of a cloud service called Prezi, which is pretty cool. The only reason I mention it is that I may have to tell my friend in the back every once in a while that I need another chart up, because this is all beta stuff. Hopefully, you'll like it.
fiscal year 2007, as it turns out, was the last time we did an analyst day, which is shocking to me that it's been 5 years. It was also a record year, and you can see what the year was like. Then fiscal year 2012 obviously was a very strong year, and some of that was just related to, obviously, flood issues, and some of it obviously just to the industry dynamic. Average capacity per drive or petabyte shift. If anything, obviously, in fiscal year 2012, we're constrained by the flood, not helped by the flood. You can see just the massive growth that's occurring in the demand for storage. Yeah, there's all sorts of different devices that need to be used in the storage hierarchy.
Clearly, at the end of the day, everything goes on a disk drive in some form or fashion, and usually multiple copies go on disk drives. You'll see the petabyte growth rate at over 40% over that 5-year period of time, which drives a revenue growth rate of over 5%, which is what we think will continue to happen. A little bit higher than the long-term forecast that WD gave. Good net income growth and cash flow and gross margin. It was a good year, obviously. Again, as you all know, we're very focused on returning value to the shareholders. I can't remember what the story was that someone was telling me once about I don't know. I was somewhere, and someone was saying something about maybe it was one of the employees. That's what it was.
It was one of the employee feedback questions. I'm looking at Bob because it was probably the engineers. How come we can't just take that money and spend more on R&D? Or another one was, why can't employees just get paid more? I had to say, "Because it's their money." There's a funny attitude by companies that sometimes I think they forget it's their money. We're big believers that obviously you need to be disciplined. You have to have enough money and manage your cash flow in a way to invest in R&D and capital to take advantage of all the opportunities that you have. If you're wildly successful at what you do, by definition, you should create more value than it costs to develop and market that technology. The question is, what do you do with excess profits?
We just believe that a lot of it should go back to the people who actually own it, and then they can decide what to do with it. We're going to keep that mentality up, and maybe we're blazing a little bit of a trail as technology companies slowly start to creep down the dividend path or other creative ways of returning shareholders. We're going to keep doing it. Pat's going to talk about that a little bit later in terms of how we think about it. Company's got 55,000 employees. We're spread all over the world, as you know. A little map that shows you the different locations and what we do there. A couple of things I would like to point out on the management team, because maybe people haven't thought about this holistically other than announcement by announcement.
Of our senior leadership team, which is, say it's 20 people, really six or seven are new people to the company with broad industry experience, and it's one of the things I'm most proud of. As CEO, I think it's really important to continue to bring in new talent and to attract people from leading companies and help us think about ways that we move forward, not just in our core business, but take advantage of some of the real opportunities that are presenting themselves as storage becomes more fundamental to all of our lives as users. Same thing on the board side. We have five new directors out of the 11 or 12 directors that we have with a bunch of great experience in terms of either Asia focus. We have a board member from Samsung, Korea.
We have one who's been based in China for 35 years, chairman of Ford China and Siemens China, and he started his career at GE. It's an interesting career. I think he's worked for three of the Fortune 10 companies. Jay Geldmacher from Emerson Electric. We just have a really strong board. Bill Coleman joined us from BEA. We're getting a really good sense of some of the issues that are going around virtualization and software. I'm really proud of my board, and honestly, I think it's one of the strongest boards that I've been associated with in terms of all the different technology companies I'm aware of. I think it's a big advantage for us. Today, we'd like to just make sure that you leave with a few things.
One is, I just want to go quickly through our view of what's going on in technology and how it led to our strategy to provide a basis for your questions about what we're doing and why. The biggest thing is, we believe mobility in cloud is real. It's happening, it's happening quickly, and it's going to fundamentally change computing. There's all sorts of implications for us, for our customers, for our customers' customers, and it's basically one that is driven around more storage and different types of storage devices and hierarchies that Seagate can play a role in.
The second is that Seagate, because of its position in storage in general, and clearly in enterprise storage, we have a high degree of engagement and capabilities around working with end users or with OEMs that are addressing these new architectures to say, what can happen at the device level in order to make some of these new architectures more scalable, cheaper, or more reliable. Finally, again, we think that the financial model for the company is quite strong. On a relative basis, our capital needs for petabyte shipped are quite low, really. Whether or not it's 5% or 6%, we think of it more in terms of dollars, somewhere around $700 million per year.
When you compare that to alternative technologies like silicon, where 30% of the revenues have to go into capital, this is a pretty cheap way to provide storage, which again, at least for mass storage, is why we believe that there's such a big opportunity for us. Same thing on the R&D side. We have obviously plenty of margin to fund R&D and to position ourselves for new markets that might develop. Those would be the three things we'd like to leave with. I'm going to go through this fairly quickly because I think most of you are probably grounded in this anyhow, but just so you know where we're coming from in terms of our view of what's happening from a technical perspective, and then therefore how we get positioned. I'd like to talk quickly about consolidation.
This is an industry that's had 247 companies at the drive level that have been in existence since 1957 or whenever. That was the year I was born. 1957, whenever IBM first came up with this idea, which Shugart has a great story about, but I'll tell you that in the after story. 247 companies now consolidated down to three. That's, obviously for those of you who've been along, looks like Dan Fenton in the crowd, and Shooty have been along for the ride for 30 years or whatever. It's been an interesting ride, but it's one that we've always said, this is where it's headed because this technology is really hard, and in order to generate the R&D and capital to be successful, the biggest ones are going to win.
It's now down, obviously, to an industry structure that, again, allows us to align supply and demand, I think, much more reasonably. It also makes us, I think, much more conservative about capital deployment in response to demand as opposed to in advance of demand. I think that's one of the biggest changes that's occurred as the industry's been able to consolidate. Interestingly, also, what's happened in the upstream suppliers, I think people are less focused on this because if you look at heads, and medias, and motors, and heads and medias are clearly the two most important, there's been massive consolidation here. In fact, that consolidation occurred prior to the drive industry consolidation. The importance of that is, this is really where the core of the technology is.
Yes, I'm not trying to negate the difficulties of firmware and integrating heads and disks because it's really hard to do two and make that stuff manufacturable. The core technology is really heads and disks. As that industry has consolidated, it just means that there's less parts out there for people to take and integrate into drives. Again, so the balance around how much capacity there is or how many available products there are, what's the pricing of those technologies, has changed dramatically in the last 10 or 15 years. Really, TDK is the only external head company, as you know. A couple of media companies, even on substrates where it says there's six, there's really only one glass substrate company, which is the one that makes most of the substrates for notebook drives.
Big consolidation there along with consolidation at the drive level, again, means throughout the supply chain, supply-demand alignment is a lot easier than it was historically. I think there's going to be long-ranging implications for that. Okay. Architecturally, lots of people talk about lots of things, in my little simple world, I try to break it down to how do I explain this to myself or to friends who aren't in technology. To me, it's just cloud mobility open source. No matter what vector someone wants to talk about something, to me, it comes down to it's cloud, it's mobility, and it's open source. For those of us who've been in this industry, like I have since 1976, the cloud is time share, updated to 2012, with a lot lower cost computing and a lot better interfaces and all that kind of stuff.
It's basically the concept that says, look, if you can share an infrastructure, then it gets cheaper, and that's good for people that need to use technology. It's just doing it with different tools now that are a lot easier to use. They're more scalable, meaning that if you want to add infrastructure, you can do it more incrementally instead of massive hunks, which again means you can address computing needs easier. You don't have to get in front of it with big investments that then you have to hope it gets filled up. Leveragable, meaning that once you have that investment, you can do a lot of things in terms of different applications that can run on that infrastructure. Of course, available, meaning that it's always on. Lots of implications here about what does that mean for storage that we'll get to at the end.
Mobile clearly means connected, social, democratic. I just view it as the more people that we can get talking and sharing and creating information and sending it all over the world, that's a good thing for a lot of reasons. We've seen some of the positive benefits of that in terms of some of the social changes that have occurred in the world, but clearly it provides for a richer human experience and, notwithstanding maybe the bad map offering on the new Apple phone that, at the end of the day, drives more storage up and down the pipe because that end device needs to have fast response time. Bandwidth is expensive, bandwidth is slow, and therefore caching has to occur. It's not just that all this stuff has to be stored somewhere as an endpoint.
I think it's lost on people that this stuff actually has to be stored up and down the pipe as well to get the user experience that you really want. You don't want to deal with latency issues. Then open source, I think is the other big one. There's so much effort that is going into looking at the gross margin stacking that happens between the device offering and the end-user purchase, that new technologies and new companies are basically going in and saying, how do I disaggregate that software and services stack and offer it much cheaper than it's being offered by a lot of the big companies today? They're doing this obviously through open source platforms.
The implications for us and the opportunity for all of us as computer users is really astounding because, again, it means that we can think about architectures that are much less expensive to deploy, which means more applications clear IRR hurdle rates, which means more projects get approved, which means more devices get sold. If you're a device company, anything that's going on in the world that breaks the technology hold that other companies have that effectively constrain the deployment of your device is a good thing. A lot of people talk about the commoditization of storage, and they, I think as investors, sometimes it gets confused about, oh, that must be bad for Seagate. Far from it. It's not the disk drive they're commoditizing. It's the fact that someone buys our drive for $150 and turns around and sells that for $15,000.
They sell that because there's a software and services stack above it. The question is that software and services stack worth $15,000? The answer is, in many cases, no, because you can see the gross margins that those companies have. There's hundreds of billions of dollars of gross margins that's a jump ball that a lot of companies are after. As they bring that down, it obviously means more devices get sold. We think this is a big deal, and you'll see Seagate making a lot of investments in technologies that drive this trend towards lowering the cost of computing. Another big thing, so that's kind of, if you will, architectures. The other thing I think about is what I call who's using this stuff and the convergence of consumer and commercial markets.
Again, if you look at the success of Apple, I think one of the real lessons there is clearly that if you can be successful in the consumer space, it has huge implications for you on the commercial space. I would argue that it's the same thing, that if you're successful in the commercial space, it has potentially huge implications on the consumer space. Why? Because people are using the same infrastructures and architectures. It's the same UI. It's the same experience. Therefore, as a company, we need to think about what are we going to do from a product perspective that allows us to leverage our commercial expertise and market share to transition into being just as effective in the consumer space.
That's why we bought LaCie because obviously that company was the most successful company for 30 years in being effective in consumer deployment, and we're doing a lot of other investments just to make sure that as people use more end devices like iPads and smartphones that are storage-like, that they therefore can leverage off of a storage infrastructure that can enable and enhance that experience of being basically storage constrained. The third is what I call applications, and I think this is where people get very confused about the role storage plays or the role that HDDs play. I break it into you're either a consumption, it's a consumption application. I'm just looking at pretty pictures. I'm a knowledge worker or I'm a creator, whether or not that's PowerPoint or spreadsheets or video shows, or if I'm just making sure my stuff is safe.
Depending on what that device is, yeah, there's different storage needs. If it's just a consumption device, it doesn't need a lot of storage. There's got to be a lot of storage somewhere to feed that content or to create that content, but to watch that content, yeah, no, flash is probably fine. If all we are is a world that watches stuff, I would worry. Thankfully, we're not a world that just watches stuff. We're a world that actually creates stuff every day, and we create value in lots of different ways, and we need lots of powerful tools to do that. The percentage of those devices that needs disk drives or other types of mass storage is high. Obviously on the storage end, it's all rotating storage or some element of flash for acceleration.
I think about from a strategy perspective, we think about what are the products and strategies that we need depending on if we're addressing devices that go into a consumption market, a creation market, or a pure storage market. They all feed off one another. The other thing is infrastructure. What is all this stuff running on? Hardware, processors, networking, and storage. Interesting dynamics between those three pieces of hardware. The processors, up until recently, I would say, are fairly standardized, but that was really a virtue of monopoly. Then really until the advent of Arm and until the advent of companies like Qualcomm that came in from a different perspective on the communications side, the processor world was a pretty good world to be in.
I think it's obviously going to be a more challenging world for the big processor companies because there are all these competing technologies. Networking, similarly, the devices that are networking have been fairly proprietary in their nature, and that's probably going to be challenged. Storage has been a fairly open, standard-based technology for a long time. I think it's just one of those great examples of oftentimes what's your greatest weakness becomes your greatest strength. I think from an investor perspective, historically, people are negative on the drive industry because "it's a commodity." I think ultimately this is going to be our advantage because no one's out trying to take our margin because there's just not enough margin to make it worth the billions of dollars of investment and thousands of engineers that you'd have to hire.
The result of that, I think, is as these technologies get deployed that start breaking down some of these blocks on processor networks, again, it means a higher degree of proliferation of storage devices, which is going to benefit our industry. On the software side, it's all about security, ease of use, availability, analytics, big data. We can throw all the fancy words in there. It's that relationship between what's the software doing and what's it doing with that hardware architecture. Look, those have to be coupled for optimization. Another big lesson of Apple is if you're in the device business and the software business, you likely make a better product than if you're not.
It's because you can do all sorts of things between that layer of software and hardware to optimize the solution, whether or not that's a cost optimization, a user experience optimization, or a reliability optimization. The funniest one I love is virtualization because investors will throw virtual, but it's going to be virtualized. In their minds or then two derivatives down, if you talk to someone on the street, somehow they think that means there's going to be no hardware, which is really kind of funny because where is this code going to reside exactly? It all sits on hardware devices. It's funny to me that people forget that there's something called machine code, and that's the layer that matters, right?
It's that layer that says, if you have a software, a higher level, abstracted level of software running that's trying to make a machine do something, it's got to go through something called machine code, but it still resides on machines. At the end of the day, it's this interrelationship between hardware and software that has to be reoptimized for this more scalable, leverageable, available platform. The engagement that Seagate has with leading software companies or leading systems companies that are trying to, and those can be big companies or small companies, that are trying to solve this problem is very high because people understand that it's the way you make that hardware device respond to this new level of software that's going to break through some of these log jams that we have or some of these control points that we have.
This is, I think, a big dynamic that we're spending a lot of time working on. I'm going to run through this. Let's just talk a little bit about the disconnect between petabyte growth, areal density growth, what that means for our industry. Clearly, again, as the world is deploying these new architectures and as mobile devices come online and generate more data flow around the system, you see this acceleration in petabyte shift. This is just a graph of both the annual shipments and the installed base of HDD. A couple of little timelines there to denote when smartphones kind of made their appearance.
Obviously, smartphones are approaching 1 billion units a year here in the next year or two, probably. Tablets is that next thing in 2008, 2009. What you actually notice is the curve actually starts getting steeper as these devices get deployed. It actually doesn't get flatter. That's because more people are getting connected.
They're sharing more data. They want to see richer content, that drives end-user storage, and that drives HDD storage. That's a trend that we think continues. This shows a 57% growth rate. If I think back 4 or 5 years ago, petabyte growth rates were in the mid-20s. Areal density growth rates were probably in the mid-30s. The industry could easily meet that demand by just staying up with areal density and not adding heads and disks. About 3 years ago, that shifted where petabyte growth rates kind of kicked into the mid-30s to low 40s. Areal density started going down as we got to the harder end of our technology deployment on the current generation of technology. Petabyte growth rate went to 50%. This year, it'll probably be over 60%.
In large data centers, it's over 100%, areal density growth rate is now down to 25%. That disconnect is something that's fundamental to either believe or not believe, understand or not understand. If you think about it basically says that the demand for storage is growing at twice the rate of the industry's ability to provide storage. Therefore, without substantial investment in heads and disks, which is what it's really all about, there's going to be, I think, another sustained shortage. These are the challenges we think about, which is how do we manage our investments and how do we deploy our capital to meet those trends. Here's the areal density trends. You can see that the perpendicular. These are mostly demos, just so you know. These are demo rates, with perpendicular, obviously, we've been shipping perpendicular for a while.
You can see that perpendicular curve rolling over. That's when areal density started slowing down from those mid-30s to low 20s to maybe even sub-20 right now. Shingled recording, Bob can talk about this later, is now being demonstrated at areal densities higher than perpendicular. We'll ship our first SMR drive this year, we'll be the first in the industry to do that, I think. At the meeting last week, I think they said something about in a couple of years or something like that. We're excited about where we are with shingled recording. We think that'll give us an areal density group boost. HAMR, Heat-Assisted Magnetic Recording. For those of you familiar, I don't need to explain. For those of you not familiar, it's basically putting a laser on top of a head.
The head's about as big as a little piece of pepper, heating up the media really quickly, writing, and then moving on. It's really hard. We're feeling pretty good about where our technology is, and that's going to be another kick to areal density. If we execute to all of those, that still only gets us to an areal density growth of about 25%. I think these are the types of investments that we obviously have to keep making, but these are the trends that we're on right now. That's the disconnect. Other big thing is where are things being stored? This gets Rick to your question. Is it secular? What's going on? Is it tablets? Is it this? Here we have a little bit of a where HDDs were shipped between tablets, client, desktop, cloud, and enterprise.
You can see the colored things. It's about 450 exabytes, which by the way, is a lot of storage. Just to give you a scale of what is a 450 exabytes. If you're familiar with the flash world, which we get compared to a lot, Samsung's FAB 16, which is their latest state-of-the-art, highest density, best use of capital FAB, can probably output seven exabytes of data a year. That was a $12 billion FAB. With a trillion dollars of investment, you can start to make an inroad on what we ship every year. No one signed up for a trillion dollars of FAB investments that I'm aware of. Here's what gets really interesting, though. By 2020, that number becomes seven zettabytes. We're at the low end of what we think is going to be demanded. Companies like SanDisk would say 12 to 15 zettabytes.
EMC says 35 zettabytes. Yes, there's a shift of where that's going to reside. There's going to be a big shift on an overall percentage basis. We're going to go from about 25% of that storage being "in the cloud" or basically in timeshare, and it's going to move to 60%. That's a big change, it means lots of high-capacity drives. It means lots of heads and disks. That's good for Seagate. What's also interesting is look how big the client pie is. We did these to scale. Yeah, the percentage that goes to client is smaller as part of the total, but it's still a massive amount of petabytes that have to be shipped into the client. These aren't markets that we can ignore.
We just have to decide how we, again, balance the constrained petabyte growth that we're going to have in terms of areal density. Okay. Our vision. Based on those theses, cloud, mobility, open source, consumer, commercial, underlying architecture is changing, we came up with our vision and strategy. This is the exercise that we went through over the last six months. Our vision, which is every company should have a vision statement, is very similar to what it's always been, which is to lead the world in storage, protecting, and sharing its digital creations. We think that's a good mission. We think that's an important place to be in the world. One of the things I'm proudest about in terms of, Not proudest, most honored about in terms of being CEO is I think what we do is really, really important.
I think we help make the world a better place because our technology's been deployed and making schools better and making medicine better and all sorts of things, making financial markets more efficient. There's something about going to work every day and knowing that what you do is actually important. We try and make sure that that's reflected in everything we do, and we take it seriously. We're deep in innovation and leadership. Seagate's always been about being a technology leader. We believe that innovation is something that we need to continue to do every day.
We have to think about how we continue to invest in our core business to drive areal density, we also have to be creative about what are the other opportunities that are occurring because of changing architectures and storage, what does that mean in terms of our product strategy, and to always extend our technical leadership. Just a little bit of data on that 4,000 engineers, 7,500 patents. The company has hundreds of firsts. We've just put up some of the recent firsts, this is something that you'll see us continue to invest in.
Even though we do believe that it's important to return a lot of that excess cash value to our owners, we also realize that it's very important to keep investing in our underlying technology to make sure that we can continue to do the kind of cash generation that we are doing for the next 20 years. I think it's a good balance, we're very serious about being a real technology company. What we do every day is phenomenal, people are coming up with innovations that are really mind-blowing in terms of what our technology does. As you can see, the application environment that we sell our technology in is broadening by a fair amount. Our priorities. Cloud devices and service.
We have a big focus, both technically and organizationally, in making sure that we have product strategies that are aligned to the various needs of people that are deploying clouds, whether or not they're public or private. Not just at the device level, but at the integration level between devices and software, also at the services level. We do know more about enterprise workloads than any other company in the world. We keep track of everything that happens when it happens in a disk drive. We know when we're being asked to seek, we know when we're asking to write. We watch patterns, we move data around to enhance performance.
There's a lot of things that go on in that little drive that has 6 million lines of code in it that we can leverage into a better user experience as people start to deploy these new architectures. We're very focused on this. You'll see us, for sure, have devices that are optimized for cloud environments. You'll see us, for sure, providing some level of services around cloud providers and probably even providing that service directly, if for no other reason than as a prototype to test some of our device technology. This is obviously going to be a big area of growth, I think, for a long time. I think the shift to cloud and mobility is the biggest thing that's happened in the 32 years I've been involved with technology.
I think it's bigger than client-server, and I think it's going to be more profound in terms of the opportunities it presents for a technology company. Why do I think it's bigger than client-server? Is because it impacts consumers and commercial. Obviously, the world's a lot bigger. If we go from 1.7 billion connected users to 2.5 billion connected users in the next 5 to 10 years, obviously that's a pretty special thing if you're in the business of providing the base technology to do that. End-user focus is obviously got to change as well. I think this is one of our biggest challenges. Our industry and our company has been designed to engage with a fairly limited customer set in a fairly defined way to provide technology to the world.
That architecture is changing and the customer's customer is changing, which therefore means every way we engage with our customers and other end-users is changing. Organizationally, Rocky has a lot of work in terms of saying, what do we need to do in terms of our engagement with users and end-users, and therefore realignment of a sales and marketing and support infrastructure, which then, of course, feeds into the product roadmap to address these new opportunities. Not going to be easy, but if we're successful, the rewards should be pretty substantial. It's really the first time in 30 years that this transition has happened. Consumer, again, it's not just about re-engaging with big companies and big governments that are deploying these new architectures.
It's that there's now this consumer world out there that's gobbling up storage left and right, that needs devices that are easy to use, that can talk to these mobile devices, that can make sure that when you walk in your house, you know that what's on your phone is synced. You don't even need to be asked, and you don't need to plug anything in, and you don't have to say, do you want me to keep the pictures or not? This is the world that's in front of us, and it's in front of us really in the next one or two years. Again, one of the reasons we bought LaCie was to advance our opportunity set here. The operating model has to be flexible.
We have to be able to scale our manufacturing infrastructure up and down to address whatever the supply-demand environment is, in terms of whatever shape that takes in terms of number of units and gigabytes per unit. We think that we have a really, really good model. We led the way with Factory of the Future in 1997. That took us until about 2003, I think, to fully deploy. Dave will talk more about what we're doing going forward, we're on the front end of what we think is a pretty exciting transition in terms of how we manage our asset base. We're basically very flexible in terms of changing demand patterns, whether or not that means overall or whether that means how we invest in individual device technology.
At the end of the day, it's obviously all about the people you have and the technology you have. Seagate's a good company to work for. We pay a lot of attention to our people. We try and stay on the leading edge of lots of things around making it a good workplace. We're responsible to the communities that we work in. We're very responsible to the environment. Obviously, we've always been built on a core of technology, and we're going to continue to do that. That's the vision and strategies that we have. Again, just in terms of takeaways, we do believe in mobility and cloud transition. Biggest thing I think that we've seen, and we think that creates a lot of opportunities on the storage front.
We're going to use our technology and our engagement with our customers at the technical level to address the needs that they see as a result of these changing architectures. We're going to continue to focus on making sure that we have a financial model that not only allows us to fund those opportunities, but also to return substantial shareholder value back to our owners. With that, I'm going to turn it over to Q&A for a second or 20 minutes or 30 minutes. Before I do that, I'm going to ask the man behind the curtain, John, to come out front because we have something to show you. He's got to unplug something. The Wizard of Oz. We've been running this whole presentation on a HAMR drive. He is now unplugging the HAMR drive and bringing it out there.
John, will you please verify for the crowd that that in fact was the drive that we were running the presentation off of?
I verify that's the drive we've been presenting off of. One of the first four HAMR drives that we've manufactured. It was manufactured by hand during the process of automation through our manufacturing process. Having the drive available allows us to iron out the viability of the components and test them in the platform. A lot more convenient than something that we can manufacture in far larger volumes.
This is what, a terabit per square inch or something like that? Is that what we're?
This is actually an enterprise drive. We run it below what the capability of the head is on the spin stand. It's 500 gigabits per square inch. As I'm sure many of you have heard, we're at a terabit per square inch on the spin stand. For enterprise drives, they just follow at a lower gigabit.
Yep. Great. Thank you. Yeah. Look, that's exciting. Pat's going to be running his presentation on something else really exciting. We'll talk about that in a second. Yeah, it's a neat piece of technology. The fact that we have it in form factor and running and working, we're very encouraged about that technology moving forward over the next 2 to 3 years. All right. Q&A? Q&A. Yes. Well, Bob's always up for hiring more engineers. The question is, I'll try and recap these questions for those of you in the back. The question is, if the secular trends are that there's less clients, more enterprise, what does that mean for your business model, your engineering talent, et cetera? You said something that I want to clarify, that if the client PC side shrinks, this is the same characterization that people had about desktop and notebook.
It's on a relative basis it's shrinking. It's not shrinking. It's still growing. The question is it growing as fast as it used to? When people saw notebooks accelerating four or five years ago, the mindset of investors was that desktops was going like this and notebooks were going like that. That isn't at all what was happening. Notebooks were growing like this and desktops were growing like this and they started going like this, but they were still growing. In fact, desktop still grows today. I think the secular trend isn't that clients are going to go down. Maybe they do, but they certainly grow at a slower rate, which is your point anyhow.
You kind of saw, yes, we believe that that shift is happening between where petabytes are shipped to, that more petabytes are going to ship to "enterprise cloud" on a relative basis than to client. You also saw the size of those pies. The number of petabytes that are getting shipped in the client is still 20 times what it is today. There's still a lot of single disk, two-headed devices that got to get made, right? In terms of margin, yeah, you're right. That's what I was saying, that if we're shipping more heads than disk per spindle, that's a good thing for Seagate being a vertically integrated company where most of our R&D and capital is in media and heads, because then it means we get more absorption off of those factories on a relative basis.
Because the easiest capital to change is final assembly and test. Look, if you had to paint a world to change, it's probably the world we would want to paint because it's the easiest asset for us to manage once we become committed and have a better understanding of what the timing of those trends are. From an engineering perspective, I don't think it changes anything. At least not in terms of areal density, not in terms of heads and disk science. We're driving that as hard as we can. Look, the mistake Seagate made in 2006 was we shifted our emphasis on driving areal density growth as hard as we could and focused more on broadening the product portfolio. We were using the same amount of R&D dollars, I think, in a certain way, but there was a little bit of a shift in what the emphasis was.
When the management team changed, we all agreed we're never making that mistake again. You have to drive areal density as hard as you can. You can always decide how you use it. If it happens to be that you have more areal density than you need, great. You can get more margin by flying the head higher or whatever it is. The reality is the areal density is behind the petabyte growth rate, so we have to drive areal density as hard as we can. We're doing that today. Would more people help that? No. I think that we have the number of people that we need working on the core technologies. Long-term, I do have concerns around where do we find that engineering talent, and is that correlated to where our engineering talent is today? The answer to that question is probably no.
As a long-term challenge, one of the things that Bob and Dave and I spend a lot of time thinking about is, Mark Gray, who heads in this for us, where do we have our technical investment to make sure that we have access to the engineering talents that we're going to need in 10, 20 years? Because it's probably, unfortunately, not going to be in the U.S. to the same degree that it is today. We're very fortunate. We opened a fab in Northern Ireland 10 years ago now, right, or more, I guess. That originally was done for production only. We had a lot of incentives from the government to do that, and there's a good technical talents base there in terms of engineering. We also developed these really deep relationships with the universities there.
Now we're doing a lot of our head R&D, probably about 25% of our head R&D there, and we're shifting so that we have an even stronger technical capability in terms of wafer-level head design in Northern Ireland. Obviously, the biggest place that you need ultimately exposure to is Asia, given the number of engineers that are being produced there. We're doing that in certain countries. There's the ability to do that in terms of attracting talent, but it might be expensive. In others, you have to worry about IP. You have to be very thoughtful about what technologies you're developing and making sure that the property rights laws are evolved in a way that it's worth the risk to develop there. Yeah, those are things that we think about. In terms of level investments, no, I think we're at the right level. Yes?
Yeah, thanks. Scott Craig, Bank of America Merrill Lynch. Can you discuss in the five-year CAGR of the 5.5% revenue growth, what sort of assumptions you're making for penetration of hybrids into thin and light or Ultrabooks or whatever you want to call it, and how we should think about the roadmap or goalposts over the next year or so to show that you're getting traction there? Thanks.
Well, I don't know that over the next year has anything to do with what I was showing there. When I say long-term, I think five and 10 years, not next year. I think in terms of hybrid, I'll let Rocky and Bob talk to this in a second. Look, my thing on hybrid is that within five years, 80% of the drives that we build will be hybrid. That they'll be offered that way. I feel like we're Toyota. If you want to buy a hybrid, we have it. You want to buy a gas engine, we have it. There's lots of reasons why hybrid should be an attractive technology, mostly around performance. You can split that market into the client side and into the enterprise side. In both markets, there are significant advantages to hybrid technology when you think about performance and cost.
On the client side, it has to do with the user experience that people now are accustomed to because of the iPad, of relatively instant on and certainly instant refresh on warm boot. We can do that today. Our cold boot times are within two seconds, and warm boot is actually faster if you are running a bunch of applications. Bob can talk about why that is. Yes, I think that hybrid will be deployed into thin and light, and I think it will be the majority of what people are buying in two or three years. The benchmark I would look at is probably if you force me into saying, where would I like to be in one year? If we are selling 5 million hybrid drives into the notebook market in 2013, I think that would be pretty good.
There is a couple of OEMs who are definitely on that path. Dell allowed us to make a statement here today that basically they are going to buy hybrid drives from us across the entire portfolio. It is happening. Obviously, as they gain market share as a result of the performance that is attributed to that technology, it will obviously force other companies to respond. On the enterprise side, it is more about the $ per IOPS, but we think that the story is just as compelling, that there is a lot of reason why hybrid drive is important. Yeah, all of that is thought about in terms of-- My point is, I think that the petabyte growth rate equals or exceeds what we have seen the last five years.
In my mind, since all we do is really ship petabytes and we try and do that as attractive a cost as possible, then our revenue growth rate will be equal to or greater than what it has been. Do you guys want to talk a little bit about the market dynamics around hybrid deployment or the technology? Rocky, you want to first talk about that?
Sure. I think that the hybrid and the client side is going to make a big difference with the reinvigoration of the notebook-class product. I think the hybrid provides a number of benefits. One, cost. We are going to be able to enable the $499-$599 notebook thin and light-type packaging, which we do not see today with the pure SSD solution. It increases reliability, and it enhances power management and therefore extends battery life. If we continue to proliferate, the hybrid solution should be pretty compelling to sustain and actually reinvigorate the client growth rate, I think so.
I guess the only thing I'd add, just from a deployment execution Steve talked about across the portfolio. This year, we've got our third generation notebook. We've been working on the technology for a long time. Involves a lot of firmware, integrating features into a controller, there's silicon properties that go along with it. We're well down the path, and it allows us to do some things like get the least amount of flash on the drive, and get that experience for the consumer with the least amount of cost. We feel good about that. Intel is doing, as you guys all know, their Ultrabook is going to be big next year. We've been developing with them. We're the only one that's been developing with them for the last nine months, that's been a big engagement for us, kind of a unique experience.
This year, a couple of things, and Steve mentioned it, are going to happen. We're going to put that whole technology on desktop, we'll start shipping that product at the end of this calendar year, just within the next couple of months. The same on enterprise. Totally different equation. As Steve said, we'll double the IOPS of our enterprise drive. If you look at our mission-critical, it's been years since we've had a big performance bump. We make incremental improvements, going from 10K to 15K, for example, in rotation speed, was a big deal, and this allows us to double it. We think it's going to be a breakthrough year in terms of just getting that portfolio coverage.
For the enterprise customers that we've shown this to, I think this is accurate or pretty darn close. Every one of them so far has asked for exclusivity. It's compelling because a lot of the solution sets that our enterprise customers are trying to solve are around $ per IOPS, this is a big issue. Bob, just one thing, Dave, while he's answering, I want you to think about the absorption question, how you think about running the factory and what happens when demand changes quickly on short-term and also long-term, how do you think about managing that deployment? On hybrid, because I think it's important, especially as there'll be more competitive announcements around hybrid. Bob, just a touch more on what is that technology.
I think people think, well, you just take some flash, you put it between the DRAM and the spindle, now you have a hybrid drive. Look, if it was that easy, it wouldn't have taken us three years or four years or five years to figure it out. Maybe just a little.
Yeah. I think the best way to think about it maybe is what people have done with flash cache modules, which is out there in the industry, and it provides an advantage, but you also have to spend two or three times more on the flash than what we have to do when we integrate it together. It gives you a bunch of advantages in terms of not only cost, but also performance. We can, what we call self-pin applications right at the drive level and not have it above the device level. It gives us, again, a cost advantage and also a performance advantage. It also is good for the OEM because you don't separate the data. You don't have data residing on two forms of devices within the unit. It gives us a bunch of advantages.
There's also, a lot of people always compare performance between silicon and HDD and try and make it sound like every application's the same. Clearly, there's differences between whether or not it's a random seek or whether or not it's a sequential read or write. When you actually match sequential performance, HDDs are pretty darn close and certainly well worth the one-tenth the cost that they are versus flash, which is, by the way, still the same differential. What people don't realize is even in the random world, what appears random really isn't, oftentimes. A lot of what we do is we watch random reads and writes, and we look for patterns that then allow us to reorganize the data into a sequential data set.
This is a lot of really smart people working for a whole bunch of years on algorithms to do that. That's a lot of what's going on behind what happens in terms of why is the performance of a hybrid drive actually better than a pure SSD when you actually know what the application set that you're dealing with. If you're a user that's going from the internet to a spreadsheet to a PowerPoint, and we basically start learning a lot of those trends, we actually can queue up a lot of things on the rotating media so it could pump through the silicon much faster than if it was just on the silicon itself. For those who've been to CES, you've seen demonstrations of that technology. This is what's appealing to our customers.
Dave, not to bring up the question, it's probably out there. It was embedded in the earlier question. How do you think about the Factory of the Future? What do you think the big challenges are around if demand is X or Y? What happens when it changes in the short term? What if it changes over time? How do you think about it from the ops side?
I think from the upside chase perspective, that's the easiest thing to answer first. We did a lot of drives last quarter, and we have the ability to go back up to that level or higher if we need to, in all of the various factories. That's lead time associated with upsides right now. I'm not really too worried about realistic upside. It's the downside case that we have to concern ourselves with. We have so many different kinds of factories. We have a wafer fab where the lead times for where the product is in the fab for longer than 10 weeks. That really touches many different pieces of equipment. That fab, if it runs lower than planned, then we have absorption issues there. There's other different operations.
I think the key point when it comes to box count at the end, though, is that the drive factories are actually very flexible
In upside or downside, because it largely comes down to the test equipment, and a lot of that's paid for. The clean room spaces, which are fully depreciated. I think we can be flexible there. Media operations, there, the media, as this petabyte growth we're talking about accelerates, media's going to be constrained. I'm not really that worried about that one. Okay. Another question from the crowd. Yep.
Thanks a lot. Ananda Baruah , Brean Capital. Steve, just wanted to get your thoughts on in a few years when you see the shortages that you talked about coming, the supply versus demand from a petabyte perspective, how do you think that manifests itself on the industry, and what do you think the ramifications of that might be?
The question is, if this kind of disconnect between petabyte growth and areal density growth sustains itself for the next few years, which I certainly think if anything gets worse because I think content gets richer, more people get connected, more devices like tablets get out there and just creates this frenzy of people sharing and creating data. Let alone what is going on in corporate world, which is more data being stored, analytics tools getting better, which means people want to store more things. Of course, regulatory reasons why you have to store more and more things. I think the big difference is that the industry is not going to lean into investments for a couple of reasons. One is the upstream suppliers will not. I showed you the consolidation in the supply base upstream, especially in heads and disk.
Those are Japanese companies for the most part, if not entirely, who are conservative on capital, who are capital constrained in a certain macro sense. Therefore, they are not going to deploy ahead of a demand curve. I think they are always going to wait and make sure that the demand is there. In fact, the reality is in a certain way, there is almost a shortage in notebook not that long ago, even with the slowness in demand because the main glass substrate provider, they were more concerned about the macro trends, and they actually held back production, and there was almost a shortage in glass substrate not that long ago, like two months ago. Right. I think there is that dynamic, that upstream.
At the drive level, for sure, it is just the balance between how you invest in R&D and capital is being watched very closely in terms of what that return profile is. I think if demand accelerates, as Dave just said, we always push ourselves to get more units out of the capital deployed that we have now versus saying, "Oh, we need to build a new factory." There is no scenario that we see that says you build more factory. It is just like, no, you get better yield or you do things more flexibly or you incrementally add. There are certain things that you can do that are beyond the days of, oh, I need to go greenfield something. If demand is lower, then we cut capital. We are going to probably cut aggressively. It has already happened this year.
Our capital budgets are down substantially from where they were just a couple of months ago. I think you are just going to always have this issue in the time being that says, look, the industry has to have a gross margin level that says we can invest in R&D and capital in a sustainable way. By definition, that was not 18%-22% or whatever it used to be. That is why the industry ended up consolidating the way it did. I think the industry is going to be much more focused on saying, I got to keep these investment levels up, which means I need a certain gross margin model. You are going to run into this disconnect, and whether or not it is 2016 or 2017, it seems to me it is somewhere around there where you just do not have enough heads and disks.
What ends up happening is you have sustained shortage, which means margins will go up. Once those are more established, people will deploy capital again. You'll find a new balance point. That's 5 years in the future, but if we stay on the path that we are, that's what's likely to happen, and the only thing that would change that is if someone said, "I'm going to go build the freeway before I think there's any cars on it." Those days are well behind us just in terms of as humanity. There's just not that much capital to go do that anymore. I view it more like L.A.
We build a freeway, and the next day it's full, and people say, "Okay, I guess we got to build five more freeways." Could I ever see a new entrant to the HDD business? Only if they buy an existing HDD company. That would be not new, but transferred. No. Only because the underlying technology is so hard. You need 5,000 magnetics engineers. Where are you going to get those people? It's very, very difficult. The capital is a really tricky technology. Today, like we saw in this HAMR technology, when you think about this, you're mounting a laser on top of something that's as small as a grain of pepper, and then you're timing when it goes on exactly when you need to write something.
The media has got to be hard enough that it can take that, but it's got to be soft enough that it's kind of crazy technology. The heads now are planing in this lube. They're not really flying anymore. They're more like speedboats on top, and if the lube's too thick, it slows the head down. If it's too thin, it creates too many waves. There's aerodynamics inside of the drives. It's a really, really hard technology. There's millions of lines of code, so the firmware side of it is very tricky. I don't see so. The industry is kind of balanced pretty well right now. We have a very strong relationship with TDK, we think that that's necessary in order to manage the capital and R&D investment to get HAMR where the industry needs it to get.
We have a really strong relationship with LSI and Marvell because we think that relationship of what happens on the VLSI ASIC side is really important to what we need to deliver for our devices. I don't see it fundamentally changing now in terms of someone decides tomorrow that, "Wow, look at this huge disconnect. Let's get in the drive business." That would take years and billions of dollars, I just think it's a tough call. I think it's a really tough call. I'm not saying that someone shouldn't do it. I just don't think anyone has the five or 10-year horizon that would say, "At the end of 10 years, I could see myself making money on that investment.
With the technology transitions, I think Wall Street sort of approached these transitions with a level of anxiety in the past. So going to HAMR or Shingled, is there a chance that this is a problem for the industry? Why are you so confident that it will happen as you sort of laid out? What's the chances of maybe a financial pressure or maybe a financial opportunity as that
I don't know that Wall Street has anxiety about technology transitions. I think, for our industry, Wall Street usually likes technology transitions because that's when our margins get better. Is there a risk to the deployment? I don't think there's risk to the deployment. We've been doing this for 30 years. Look, we're running a HAMR drive today, but these things won't be in production in mass volumes probably for three years. I think we're pretty smart about making sure that the technology is stable. In fact, I think it's probably one of the biggest strengths of the drive industry, especially relative to the silicon industry. Our customers know that we do workload testing based on 30 years of experience, and we're the only ones with the datasets to do that.
Bob's got a huge lab in Minnesota that all it does is run drives on all sorts of enterprise systems to make sure that the reliability is there. Whether or not that's dealing with Shingled or with HAMR or with anything else. I'm not worried about that. I'm worried about it only in the sense of, if it doesn't get delivered, if it isn't on time, if it's not stable, then our areal density growth rate's even lower, then we have a bigger disconnect between the demand side and the supply side. I don't know, Bob, do you want to talk about how we manage technology deployment?
Yeah. I think the only thing I'd add is that we go through a recording technology transition about roughly every eight to 10 years, and we're kind of at that point right now with conventional perpendicular recording. Steve showed the chart. What we do is we start 10 years in advance on the next technology. For example, Shingled Magnetic Recording, which is just really what I'd call a firmware enhancement of perpendicular, we've been working on for over five years. We're going to start to ship that technology this calendar year as well. There's a lot of preparation that goes, and we've done it over and over again. HAMR, we've been working on over 10 years. I guess the only point is that, A, we expect to make a transition every eight to 10 years. We resource it accordingly and then go through the transition.
There's always risk. There's timing risk, but we think we're pretty well-staged.
Capital I think is an interesting question. For sure, I think the HAMR capital issues have to be better understood. It is a highly leverageable technology. It's not like Bit Patterned Media, where you'd basically have to redeploy your entire media equipment set, which would be $billions and $billions, which is one of the reasons we're not big believers in Bit Patterned Media. HAMR's leverageable. It's definitely leveraged off the existing capital base, but it's going to take incremental capital. You have to time that with the markets that will pay for it and need it, and obviously, enterprise is probably at the leading end of that. Yeah, those things have to be managed. I think at the end of the day, the technology transitions do a couple of things. They drive areal density, and they lower cost.
From a margin perspective, they're actually usually pretty good things once you get through that initial curve. Okay. Yep.
Thanks. Rob Cihra, Evercore. When you go back to the Samsung deal, part of it at the time seemed to be the go-forward relationship in terms of sourcing NAND and your jointly developed controller or tech roadmap, whatever. Has that played through? Is there a relationship there that's long-term in nature, or have we kind of gotten to the point where okay, that's done, and now in terms of NAND sourcing, whatever, play with everybody?
I'll let Dave and Bob talk to a little bit. Let me just give an overview on the two points. One is kind of technical, and the other's kind of sourcing. The technical development on the flash management engine, which is what we call that chip, has gone very well. That product is close to being ready to go to market. We're really pleased with its performance. Again, we benchmark our drives and everyone else's SSDs against our HDD test bed. You might argue that that's overly conservative because a lot of people actually put drives out there in the enterprise SSD space that last weeks instead of years. We feel really good about where that technology development has gone. We are currently in discussions with Samsung about what we do next in terms of that collaboration.
Of course, this product's going to be in the market for a number of years. We feel really good about it. On the flash sourcing side, it's a great question, and it's one of the things that was, I think, not paid attention to about investors in terms of one of the big things that we did was not just align ourselves technically with Samsung, but we also aligned ourselves in terms of making sure that we had access to the best silicon, with priority at good pricing. Maybe, Bob, you can talk a little bit about the technical collaboration and Dave on the sourcing side, but we feel really good about the relationship and know it's at the beginning, it's not over.
Yeah, I don't know that there's a whole lot to add other than we've had this relationship for about a year, and through this first product offering, we're going to see huge dividends by the technology that we can draw by doing the product together. It's been a good outcome so far, and we'll see that continuing going forward.
Steve said the relationship's going well and just at the beginning. You go back to the hybrid discussion that we had earlier, and you put 8 GB or 16 GB of NAND on every hard drive or 80% of the hard drives in 5 years or those kind of volumes. Those are substantial volumes for the NAND vendors. That's the bedrock of the whole partnership.
Fairly predictable demand. One of the things that the NAND flash providers deal with is a lot of volatility around demand, and when you're plunking down $10 billion for a fab, that's a pain in the butt not to know how much capacity you're running through that. To have a stable business model that absorbs, let's just say, 15% or 20% of that fab is a really good thing, we think, and obviously, our partner does as well. No, we feel really good about it, and it's a strong relationship.
Sherri Scribner from Deutsche Bank. Steve, you laid out a clear path of growth in petabytes and zettabytes, seven zettabytes in 2020. I want to play devil's advocate because some investors will say to us, "Okay, on the PC side, you're shipping 500 gigabytes per PC, and nobody needs all that capacity. That's really a unit play." If all of that moves into the cloud, we only need a few instances of that, and we don't need everyone to have an iTunes song on each PC. Really, that growth is going to be much slower. What would you say to those people that would argue that we're not going to see that type of growth?
Well, look, that's the thesis and that's the debate. Is it seven? Is it 10? Is it 15? Is it 35? At seven, we're way at the low end, by the way. My guess is if you asked those analysts five years ago, "Was it going to go from 40 exabytes to 500?" They would've said no, just as well. I think you have to say, yes, there's a lot of technologies that are being deployed that say storage can be used more efficiently. I would argue that a lot of those have already been deployed, especially after 2008, whether or not it's sync, dedup, higher utilization of storage platforms. People used to run storage platforms at 50%, 60% utilization, then it was time for a new one. Now they're running them to 80%, 90%. A lot of that's already been done. Yeah, it's going to continue.
That'll continue, and that's why I said the funny thing about virtualization. Guess what? It still runs somewhere. In terms of a single instance, I don't think that happens in my lifetime. I hope it does because that means I live really long. Look, when I can drive from my house from Los Gatos to Cupertino and not lose a phone call, I might start to believe that there's some day that's in our future where you can stream any movie, anytime, anywhere, and it's not going to be a very poor user experience. I think we're a long way away from that. What do you do in the meantime? You cache.
The way you make that experience rich enough is that there's caching, which means that the content is closer to that end-user device, and that's going to be on a disk drive or a disk drive and flash or whatever. The other big thing, though, I think this is where the argument falls down, is the world's getting bigger. If you have 1.7 billion users today connected and that number goes to 700 or 500 because we're at the. You're not going to be worrying about Seagate. I believe we're going 0.5 billion connected, and I believe that the nature of the content that people share is going to get richer and richer. I think it's going to be full motion HD video. When you do that math, it actually says that the seven zettabytes. Yeah, it's the question, and that's what we have to watch.
Do we believe that the client world goes more to a single disk, two-headed device and there are fewer of them because? Those are all in those numbers. We also believe that the back-end storage required to feed this system of people needing and using content every. Funny, there was an article written the other day in The Wall Street Journal by, shoot. That was talking about the celebration of the disk drive. Come on, you guys read the newspaper. Was it Mike Miller?
Malone.
Huh? Malone. Yeah, Michael Malone. That's right. It was a really great article, except that then the conclusion was and drives are going away. I was almost going to write a letter to the editor thing or whatever and just say, "Well, that thing, both in the resultant reading of it got stored on its drive that second. Look, we're a long way away from that world. I would argue, and if you look at technology over the last 35 years, what tends to happen, and this is why it's great to be a hardware company, what tends to happen is you have all sorts of people creating all sorts of solutions to address these problems, and these problems are really about latency and cost. They do that with technologies that gobble up hardware, and they're successful or they're not.
If they're not, the hardware goes away and the next guy has to buy new hardware. It's not like a code base that I can take over and put in a new platform, right? It's whatever, Dropbox and Box.net and all, they go out there and they buy a bunch of disk drives and, whether or not they make those business models successful or not, will determine whether or not they stay in business or not. If they don't, guess what? All those drives were bought, the next one that starts the next business that has a business model that's maybe more successful is going to need to buy disk drives, too. There is something about that we kind of feed that advancement of how people are trying to address latency and cost of computing.
The fundamental issue is we're creating more data every day than we have the capability to create technology to store. That's the equation I look at. If that curve rolls over, either because less people use it or for some reason, less people are sharing or creating content, then yeah, then maybe that comes into balance, and then we have more of an issue around what we do. Those aren't what the trends are saying. The trends are saying, if anything, it's accelerating the other way, especially with the deployment of devices that connect more people. That's the counterargument, and that's what we have to watch. Yeah. Okay, last question. That was really good. Yeah. Now sell your time to some of these other people.
Okay, the name starts now. Three questions. One, back to the earlier question about secular cyclical. What are you looking for to try and, questions open, sort of when and how will you know secular versus cyclical? The other are really about the mix shifts. One is we haven't really heard anything about the do-it-yourself data center guys, the Facebook and the Google that are kind of marginalizing out the enterprise storage providers, if that's an opportunity for Seagate, what you're doing there. Then finally, mix shift that we've talked about, and we've talked a little bit about client versus enterprise, but we haven't heard a lot about in the retail. I think you're going to see more and more homes putting in, let's say, the home NAS system-
Yeah
where you may not have that much on your notebook or desktop-
Yeah
I personally have several terabytes on a network at home.
Yep. Okay, let me ask-
Mix and margin impact of that sort of shift.
Sure. Let me answer the second one, then remind me of the first one and the third one. Judy kept me out too late last night. No, he really didn't. The second question was, wow. I'm so excited to answer it. What?
Do-it-yourself data centers, secular versus-
Oh, yeah. No, we did answer it. Sorry. Maybe I didn't articulate it very well. That's what I was talking about in our engagement with the end user versus the traditional engagement with our customers who engage with the end user. The deployment of architectures that we're talking about, cloud and open and scalable and consumer, that's being driven by a lot of end user IT departments, in addition to traditional systems providers, in addition to new companies. They're all about dealing with that $100 billion of gross margin operating profit that's sitting out there. Yeah, we have huge efforts of engaging with, whether or not it's Microsoft or Amazon or Google or Yahoo or whoever, in understanding what needs to happen at the device level in order to make those architectures more scalable.
That's why I was saying that layer of software and the way it interacts with the hardware, the really sophisticated engineers understand that relationship, and they understand that the tie to the collaboration with the hardware device manufacturer with various levels of machine code or firmware, whatever you want to call it, is important. We have to figure, and we are, deploying engineering and business relationships with those companies and those technology providers to address what they're really trying to address, not in just today's environment. Today's environment, they may just be saying, "Hey, how do we use a bunch of SATA drives or SAS drives, or how do we do this?" They're really talking about how do they get more leverage out of the hardware in the architectures they're developing.
What's interesting is they're all developing different architectures, the leaders, and they're all kind of optimized around different things, but most of them are optimized not around storage, curiously enough. Most of them are optimized around application performance or development platforms. Why? Because everybody's driving with their rearview mirror looking at AWS. Our industry, on the downstream end, does a lot of that. You look at the leader, and then you say, "Well, what do I need to do?" Which I think is kind of silly for innovative companies. A lot of people are looking at the AWS success, and they're saying, "Okay, I need a great application and development environment." That isn't optimized for storage. It's optimized for the application experience.
We do think there's an opportunity for someone like us who understands workloads and storage architectures to say, "Well, what if you want to do a storage purpose cloud? Here's an architecture that we think is a lot better than the one that you've deployed for application experience." That's why I was saying, we'll do that on our own as a test bed and testing out some of our device technology and software layers that we have. We do have a services business. We'll leg into that with both our OEM customers and end customers. We sell directly to a lot of the customers that you mentioned.
If I didn't talk about that enough, yeah, we absolutely believe that's a huge opportunity for us, but it's also a challenge because those aren't customers that we've traditionally engaged with, so there's a lot of engineering resources. There's a lot of interesting things that you have to deal with. On the consumer side, absolutely. I think maybe, Sherri, to your point, I think one of the best hedges of if the client world changes because all these devices eliminate the need for a spinning disk in notebooks or whatever, it for sure means that stuff's got to be stored. Because no matter what, silicon's really, really expensive. Fabs that used to cost $7 billion aren't costing $3 billion now, they're actually costing $14 billion.
The next ones are going to cost $25 billion because as you go down in linear density, the equipment set gets more expensive, silicon gets more unstable. Having it be compute grade is tricky. Therefore, there's always an opportunity that says, "Yeah, I can carry some amount of this flash with me, but I need to basically hold the mass amount of my content somewhere else." Look, I think the home market, whatever you call that, a NAS device or we're struggling.
Rocky and I are spending a lot of time with the development team about what does that thing look like and what do we call it, because if you call it NAS, that kind of turns off consumers right away because like, "It sounds like I need someone to come and put it in my house and make it work." Maybe it's just something inside of your desktop computer that actually works like a NAS. Yeah, we have a lot of investment in that area.
You'll see both in terms of partnerships with companies that we've invested with or technology development that we're doing, we're developing that capability internally, I think that consumer side of the world in terms of not just DAS, not just buying a disk drive that plugs into something, but the whole environment around wireless drives, around smart NAS, to enhance the tablet experience is a huge opportunity for us. Rocky kind of led the investment in DensBits, maybe Rocky, you can talk about that a little bit in terms of this whole play.
Yeah. I think we've actually launched several initiatives around the whole client and near field storage kind of initiative, one of which was the acquisition of LaCie and their experience with first phase NAS devices for the home and the consumer application. The second one is, which Steve alluded to, is the building of our team with companies experienced like Google, Apple, Cisco, et cetera. We've got senior leadership which have also participated in some of the most exceptional user development experiences at Apple now inside of Seagate working on that whole consumer experience.
We're pretty excited between the actual products on the roadmap and the development talent that we've added to the company to really take Seagate's efforts beyond the device to another level, which we think is going to really pay off over the next three to five years in the products that we release to the marketplace around the whole consumer and home application area. Then on DensBits, this is a new technology. It's a group of ex-Intel engineers that were focused on communication processing architectures at Intel, who have approached error correction code, algorithms, et cetera, in a much different way as it relates to storage or memory management applications, which we think provides actually a unique way to build an SSD.
We're very optimistic about proving out their technology, and hopefully in about 16 to 18 months, we'll be bringing devices to the marketplace as a result of that relationship. We feel multiple compelling forces going on around our whole consumer and home applications area.
What was the first question?
How and when will you know secular versus?
Oh.
Looking forward and time frame?
Well, you're going to hate this answer. Now I'm treading closely on politics. What the heck. I think the next five years probably looks like the last five years, and that's the world we all live in, and that's what we got to manage to. The answer is the only way you would really know is if the macro issues get resolved. All of us in this room know that for the macro issues to get resolved, some really hard decisions have to be made that our political leadership is not going to make.
You're going to have periods where instead of addressing the fundamental problem, that there's been too much borrowing for unproductive projects, and therefore the people who own that debt need to take the hit, but they can't take the hit because they don't have enough equity to fund the hit, that we're going to keep doing what we're doing, which is saying, "Let's not deal with it now. Let's hope we deal with it later and things will be better." Maybe there'll be enough growth that goes at a faster rate than the problem gets worse, so that when you deal with it, then somehow that's better. I think that's absurd. I think the problem gets worse before it gets better, but central banks around the world will keep doing what they're doing.
That's just means you're going to have six great months, followed by two bad months, followed by nine good months, followed by three bad months, and it's a tough way to run a business, but I think that's the reality. I think it's going to be really, really hard to dissect what's macro and what's secular. It's going to be hard. Therefore we got to watch our capital. What are we going to do? We're going to be really conservative on capital. Because like Dave said, if we're wrong and we're short, okay, that'll be reflected in pricing, and then we chase upsides and if we're right, then at least we haven't over-deployed capital.
That's maybe not a great answer for a growing world, but I think unfortunately, until our political leadership decides they want to put a better model out there that allows everyone to invest with confidence that the structural issues are addressed, that's what big, smart businesses are going to do. That's what we're going to do. It's going to be really hard. I wish I had a good answer. I wish I could tell you, I was talking earlier. The way I feel is, you know when you feel good, if you feel good physically for a while, you kind of forget how good it feels to feel good until you get the flu or you get a cold or whatever. After that, you're like, man, when you feel good, you really feel good, right? You don't know it sometimes till you're sick.
For a lot of my CEO pals or these guys, it's kind of like we've had a five-year flu. Now it's the new norm, I guess. One of these days it'd be great to wake up and say, "Yeah, next year it's going to grow 5%, and the year after that it's going to be seven." Man, if we get back to that world, it's going to be really fun to run a company. Because it's tough right now. It's tough for all of you. How do you invest? It's, I invest now because QE3 started, but fundamentally, I know nothing really changed and I don't know. Maybe silliness. Okay. There we go.
That's great. Thank you.
End of political statement. All right, on that note.
We're going to take a break.
Okay.
We'll be back at 10:40.
Thank you.
Thank you for joining us, everyone. If you could make your way back into the main ballroom, we will begin the second half of our presentation. If you could join us in the main ballroom, we would appreciate it. Thank you.
Yeah. Is Ronnie down or something? Yeah. All right. Okay. All right, let's get seated please, if we can. If you're still talking, go outside. Which is fine. I don't really care if you go talk outside, I'm not keeping notes. Dan Benton, are you still here? Okay. Just one thing, I was thinking about Sherri's question that I should have been more particular about on the seven zettabyte thing. For our industry, just to get from where we are today to seven to 10 zettabytes, would take about two and a half to three times the investment in heads and disks that we're doing today. If it doesn't become 10, which again is at the low end of what everybody's saying anyhow, let's just say it's six, that still takes a lot more investment than what the industry's doing today on today's margin models.
I think it's a matter of degrees, in fact, I think the right question was another one that was asked, which is, you don't want to leave $billions of revenues on the table because the capacity isn't there either. That's what's going to be hard for our industry, is to say on the one hand, you don't want to over invest, for sure, you also don't want to under invest, right? Then you're leaving a lot of revenue on the table, I think it's going to be a tough balance. Just so you understand, to get to seven zettabytes means there's a big shift in the business models already to allow us to invest literally two and a half times as many heads and disks as we make today.
It's a big challenge for us, even if that growth isn't there. Okay, I'm going to introduce Patrick O'Malley, who's going to go through some of the financial modeling and thoughts on capital allocation and things like that, then we'll open up for another Q&A.
Thanks. Do we have Nicole? Oh, here. Thanks, Steve. I'm already behind schedule. Everyone's behind schedule. Don't follow me. That's true. The good news is, it's nice to see a lot of familiar faces, so many of you have been with us a long time. You know the industry pretty well. I'm going to try to leverage a financial model of what we've seen over the last 10+ years, and how we see what Steve described to everyone for the medium to long term. As Steve said, yes, could things change in the short term? Yes, that's called running a business. We have to run our business every day. You heard some view of it with Dave and Bob of how you have to react and sense and respond, even with supply chain.
This is really for a long-term model, some of these, from time to time, will go above the range, will go below the range. Long term, we feel fairly convicted on these. Like Steve said, with just getting to the fundamental premise of where storage is growing at a rate of what we can deliver, it's not just what we can deliver, what the storage industry, whether you want to put flash or other technology could deliver, is a lagging performance. That means we're just going to be in this mode where we're always pushing that envelope to stay up with it. That's the good news. If that premise were to change, whether it's the secular discussions or whether you start thinking it, we don't fundamentally believe that.
What I'm going to go over is basically, I said, leveraging what Steve talked about, because this is how we believe it, this is what we think it results into our financial models. We sort of like this little fictitious city that Prezi created for us. It looks like the alien coming in over it. Hopefully it's not that scary because I think we can walk you through this. Like I said, many of you have been very familiar with our story, I think you could come to some of your own conclusions or, certainly we could probe them with you at a later time further. I want to open up with the, first of all, the theme, the consolidated market.
I think last year the headlines clearly were directed with the Thailand unfortunate situation with the floods, which created a supply disruption. What really was in the backdrop there was also we had a consolidated market going on to three players. As you can see, going about that same time is when the consolidation started occurring. You saw a revenue bump and a margin bump. In fact, fiscal 2012 obviously was a record year for us and probably the industry, for sure, that we've seen. There's a few things that as the consolidation happens, if I took this chart even back further, what you see is the lows get higher and the highs get higher. Some of you might see a little bit of the volatility in the curve, and that has a lot of impacts, whether supply is out of balance with demand.
When you have five companies or seven companies, it takes longer to get there. As Steve says, my guess is the industry is seeing a little bit of an industry rebalancing right now, it's reacting, I think, hopefully very quick. We don't always have insight to every piece of inventory out there, but I think we have a pretty good idea, and I think the industry is responding and going to address that. For the financial model targets that we look for going long term, 40%-43%. There's really two scale players in this industry. Steve showed the chart where at one time there was 247. There's really three of us left. Two of us are vertically integrated, so we have very similar models. We look alike. We sort of approach the market the same.
We want to make sure we get an adequate return on our invested capital. Then there's one player in there, Toshiba, that has to leverage off the left side of the chart that Steve told you about for their technology, whether it's heads, media, they have to leverage that. They're going to be a much smaller scale player. 40%-43%, we don't see as out of the ordinary, and obviously Steve talked about where we think our share may be when he opened this up, and it's going to be relatively in that range, and we think that could be maintained. Revenue growth, Steve also commented on, ±5%. We do believe, if you look over the last 10 years, it's grown at that rate.
In fact, we think as Steve showed with the seven zettabytes, it probably has opportunities to grow more, you'll have that trade-off. Do you not invest because you want to lag and then you leave some on the table, you run a tighter business. Long term, we certainly believe what's shaped in the last 10 years certainly should roll forward for the next 10, at least given where we are with the storage and technology misalignment. Our gross margin, 27%-32%. Our old margin that we sort of always walked to was 22%-26%. If you looked at the first decade of the century, it was pretty much right spot in the middle of it, and that was with many more companies. I don't know what the decade started with.
It didn't start with 247, it started probably close to nine drive companies and down to three. We don't think there's a reason why the model shouldn't move up. We certainly feel comfortable that that's a long-term model that we can certainly achieve. Operating expenses. A lot's been said about the gross margin, one of the things that people, I think, really should start focusing on in the business is more of operating margin. If you look at the last consolidations, the industry has gotten tremendous operating margin leverage. I know our competitor talked last week, they have different issues on the operating margin, I think long term they'll address them. At our level right now, you can see that's around $400 million. That's what we're targeting to run at. Could it be flexed down? Yeah, it could be flexed down.
Do we have the capacity to take it up for added investments? We certainly do, I think as I go through this, you can certainly see we have the capacity to do that. The important thing here is the most efficient OpEx in the business today. We're funding all projects we have that we believe we need right now. Can there be incremental investments? Absolutely. Steve talked about, do we need to be closer to consumer? Does that require a little more investment there? Yeah, we do that. That's fine. We certainly have line of sight of those, but we think we should be able to run this at about $400 million a quarter. Like I said, we think that's efficient. Like I said, as Steve talked about a resilient model, we could flex down or flex up.
Right now we don't really see much need other than to deliver the products we've talked about, I think that will be a good return. Talked about vertical integration. There's two of us that have that. From our standpoint today, you can see the capital over the last several years. Other than large investments, whether it's a media plant or a wafer plant, we've been pretty efficient. As Steve says, the semicon has to do a third of the revenue for a dollar in revenue. We've historically been the 6 to 8, we've, the last two years, been operating under that. We look at about $750 million that we sort of start the game with, some of that's maintenance capital, some of it's investments, whether we need a design center like we acquired in Korea. We need a facility for that.
At times you'll get things like that. Generally, we'll keep it in that range. We think we're very efficient with that capital. Right now we're probably in maintenance capital mode. This could probably support 70 million drives, 70 million plus or minus given the mix. We picked up 15 million with the acquisition of Samsung. I think the model is really scalable pretty well. We try to run the business about 20% ROIC as a quarterly target. I think the two vertical guys left clearly look at ROIC because that's what you have to make the decisions on how you're getting returns for that. Cash flow generation. As most folks here, the investors know that's one of the big stories with Seagate in this industry. We generate a lot of cash.
If you take a look at this chart that we put together for the last several years, a couple of things I want to point out. Obviously, fiscal 2012 was very, very strong, you can see the amount of cash we generate there. The start of the year, end of the year, we ran about the same amount of cash on the balance sheet. We deployed that, I'll get into that a little bit. The other number on this, if you take a look at fiscal 2009, that wasn't our best year, wasn't the industry's best year. It was a bad year for a lot of companies. Even at that year, you see we generated near $1 billion of cash flow. The entity itself can generate a lot of cash. In fact, the last five years, nearly $10 billion.
That has not been a problem for us. I think folks get really tied up in the earnings volatility. From a cash generation, the entity and the industry generates an awful lot, whether it's through efficient leverage of the operating expenses or what I just talked about, the capital. Everyone could do their own valuation metric, but if you look at last year's cash flow or current market valuation, we turned trading less than 4 times that. I think that's pretty good value. I mean, discounted value, but that's what the tagline on that is. I think let the industry decide what it's worth, but it generates, as I said, a significant amount of cash to fund all the things we've talked about here. We have some principles for capital deployment.
I'm going to build this one by one because I want to really focus on what Steve said. It is a disciplined process. We have a very active board that looks at our capital structure. Steve and I go over this fairly regularly. We want to look at this from many different ways, from corner cases of things got really tight to where there's opportunistic places. We go through a structured process as we go through this, and it is adaptable. We can flex down, we can flex up, but we look at this in many different ways at many different times. People always ask, "What does it really take to run the company?" We just had to run the company with cash. We like to say $1 billion to run the business with inter-quarter variations.
You probably run the business without stressing too much at $1 billion. If you start getting under that, you start wanting to watch your cash. I always, as a CFO of luxury, I like to say a little, $1 billion and five. Steve always says, "You don't need it." He's right, we don't. It always gives you some dry powder, and you could argue $2 billion on the balance sheet. We have enough cash on the balance sheet. The one thing you'll see, the cash really won't grow unless there is a need to do that. As you'll see here, we don't really have a, in our mind, a tremendous need to grow that right now. What do you do with the cash when you generate it?
First, you obviously invest in your portfolio. We talked about that through the operating expenses. Second, you deploy through your factories to build what you need to. We talked about that. We're fairly efficient. The third to us is you return to your shareholders. You do that in a few ways. You do it through dividends. I'll go through that in a moment. Share redemptions, which Seagate has both programs. Keep appropriate debt levels. Seagate likes to operate as investment grade, whether we're rated as investment grade. That's one thing, but our principles around debt inside the company are to operate at near investment grade level. Opportunity-driven M&A. We always have to look at these as a part of our tool chest. Our dividend.
As you can see here, the dividend has had two-step functions over the last two years, which was, as we did that, we went through and looked at our conviction on where we think the cash flow is going to be, and we felt very convicted on that. We took two-step functions on that over the last year plus, and over the last six quarters, we returned over a half a billion dollars to our shareholders in that form of a program. We think long term, the dividend program will become probably more structured as the industry aligns itself to those earlier targets I talked about, that we'd like to target a near 10% dividend growth stock over time. You could do your own modeling on that, but that's how we tend to look at that from what we see from our cash flows in the future.
The other part of the program is the share redemptions. Seagate has gone through a heavy period of the last two years of buying back a significant amount of shares. Since fiscal year 2010, we've returned nearly $4 billion for our shareholders through this program. As you saw that cash flow generation, near $10 billion, we're certainly returning quite a bit of that in just this program alone back to shareholders. That's about a 16% decrease in actual shares, which in some ways is somewhat more impressive, is that in there you don't see the $45 million add for the Samsung. We've actually took care. You don't see that step function going back up. We sort of made that a virtual cash transaction by going through that.
Our overhang of our options, our share awards inside the employee base, that overhang is cut in half in that same period of time, from 50 million potentially dilutive shares to around 25 million now. We've taken a lot of what I'll call headwinds into the face of the share purchases, and we've offset those, and I think we've done a fantastic job. Currently, we expect this quarter to be at 383 million shares outstanding. That's through private and public placements of buying shares back. That puts us pretty much on path, I think, by the end of this year, by December, that you get to our target that we put out there a year ago. It's a 350 million actual shares outstanding. We're going to be very close to that. We're on target for that.
In our last investor call, we talked about the target for calendar year 2014 of 250 million shares. We certainly think in our cash modeling that those are both attainable and achievable. Our long-term debt profile. One of the things that Steve said, what the employees always ask is, "Why don't we do X with it?" I always get back, "Why don't we pay off debt?" Well, our debt structural, number one, I think it's a good point of view to have some leverage. Our debt, we do run S&P credit rating at BB+. Like I said, we philosophically operate as investment grade, but it's a slightly less than $3 billion. The story with debt is many things. You really have to look at it as a 3D picture. It's not a 2D dimension. It's not $3 billion.
If I had $3 billion due tomorrow, that might be a problem. I have $3 billion, if you look at this over the next decade, it's not such a problem. There are two things on the chart on the right, is one, it's spaced out very nicely. I don't really have a wall of liquidity in any given year that I'm facing. I think, would I like to have our rates lower over time? Yeah, I would. We don't need anything structurally to do change that right now. If you see the first up debt is the 10% secured debt, and that's due for payment on May 2014. We do have some optionality to pay that this May, so depending on the cash flow, we may call that in, one, because of our rate, and two, because of security.
We could relieve the security off of our debt profile, which I like to do sooner than later, but that's on optionality. Beyond that, you don't see another debt payment all the way till 2016. There's no real large impending piece of liquidity over the next several years that we have to deal with. Our capital deployment plan, looking last year and this year, I put some of those pieces we just talked about in the chart here. Share redemptions, and dividend program, we were both into them last year. We're going to be committed to them both again in target fiscal 2013. Last year, you saw Steve chart earlier with about 85% payout of that cash flow. This next year is still going to be pretty significant at 70%.
The share redemptions gets us to the target of the $350, and the dividends supports what we have already disclosed for a dividend. We certainly could support that payout ratio in the absolute dollars very easily. I put debt retirement as an optionality. Don't need to do it. It'd be nice to do it. We'll have that decision at the end of our fiscal year. In M&A strategic, you saw last year we did the Samsung, we did other strategic investments. We always look as opportunity-dependent, nothing that we really ever comment on publicly, but we always look at it. As Steve talked about, we did acquire LaCie. That transaction hopefully will complete itself in the December quarter, but we're well down the path of having that asset, and we think it's going to be a very beneficial asset to Seagate's portfolio.
Let me just put all this, what we just talked about, the financial model assumptions, lay that out, what we just talked about, and the capital deployment assumptions. Summarizing again, revenue growth ±5%. Again, I'm not going to get into whether it's this year or next year, but I think long term, we certainly feel comfortable with that. Margins 27%-32%, you could almost look at our last time model high end is now approximately at our low end. Operating expenses, ±$400 million a quarter. For the capital deployment assumptions, I mean, capital expenditures, $750 million. If we flex it down, we will have the opportunity, but right now we've modeled that. Our dividends, I just talked about, nearly half a billion dollars. The share redemptions, which we had also in the previous chart at $1.4 billion.
I just said, okay, if we need the $314 principal, pay the principal off, we could do that, which would tell you that for what you'd need to do to support that program, we'd need to generate at least $2.8 billion in operating cash flows. That maintains cash. Remember I said we probably only need $1 billion. You could even probably say I really only need to do $1.8 billion if I want to take the cash down to $1 billion. I'm not saying that's going to be an outcome. The real point of me saying that is that no matter what scenario we look at, we think we can execute this plan I've just laid out with a high degree of certainty.
I'm just going to get to a point of wrapping up here a little bit, and I'm going to close with the slide that Steve opened up with on the takeaways. I'll just go through them one by one again, just to reiterate them. We do believe in the mobility in the cloud. That's probably the biggest thing driving that zettabyte growth. As Steve says, we modeled that down. We had to really pull that model down just to say how far can it go. Every time you add a device, as Steve charted earlier, you're going to have some pop in somewhere in the storage ecosystem. I think what you saw from Steve's charts, we're looking at it as an ecosystem. We're not looking at it as one drive goes away, and what does that mean? That means it goes away.
No, where does that storage go? Because fundamentally, the storage has to go somewhere. With our technology leadership, we're able to bring innovative products, time-to-market products, cost leadership products to the market, we're pretty proud of that, and we continue to leverage that. We're making the investments you saw from the OpEx. Everything that Steve talked about, we have it funded through our operating expenditures. If there's other opportunities, we'll deploy that funding as well. Right now, that leadership is being demonstrated through the products, the time-to-market products we're shipping on this last round, and some of the technology that we continue to demonstrate. This resilient financial model. I think most folks that know me personally, I know you've seen how Seagate, in times when it gets tough, if there is a macro, I know that's on everyone's mind.
We certainly can scale things back and try to run a business that still generates positive cash flow. With the opportunities, we'll go after those as well. I think that whether it's through our cash generation, our debt capacity, those type of things, we certainly will be in any position to take a look at any opportunity. That's our real focus there. You can see the top part about the operating model, but really shareholder value, we look at the whole balance sheet. We'll continue to look at that and how we deploy that to shareholders. You can see that we pretty much use all tools available to us. I want to make one more thing before we get to Q&A, because my little surprise is, I have a surprise, too. Steve's not the only one with show and tell.
This is a five-millimeter in form factor, standard connector SATA, fully functioning drive, and one just like this is running my presentation. It's out working. We're obviously well down that path as well.
Not out of form factor, the right connector.
Customers are seeing them.
We feel really good about our five-millimeter program.
That's my little show and tell.
Before we get to Q&A, I was sitting here thinking about something on the cash stuff, because it's an interesting philosophical difference that I think both investors and management can have about what does it mean when you decide that you should return cash to shareholders. There's kind of an archaic thesis that says, "Well, that means you have nothing better to do with the cash." I think is really odd, because it actually means quite the opposite. It actually means that you have enough confidence in what you're doing in your R&D and capital investments that you're actually going to create enough cash flow to sustain that growth that you actually don't need to hoard cash. It's just a different way of viewing your future prospects.
As we look at our future prospects, we believe this business creates a lot of cash flow, more than enough for us to address what we need to do, invest and innovate every day to create more value than the innovation investment and the go-to-market investment. I was at a conference by a bank not to be named, where there was a panel of young entrepreneurs with no experience, who one of the persons made a statement that, well, once you're making a profit, your company is done innovating, which to me is one of the most absurd statements that I've ever heard in my life, implying that companies like Apple and Google and Coca-Cola don't innovate because they make profits.
Profits are the result of what happens when you innovate to a degree that the value of what you innovated covers the cost of that development, plus the sustained development, plus go to market. If you have confidence that you can continue to do that, either because of your competitive position, your technology ownership, the growth of the market, and that translates into a business model that generates lots of cash then it's, I think, bad form to hoard that cash. One, it's a drag on ROE. Two, you can always go borrow money if you have some big, important need that is important for strategic reasons, and if you can't justify that to the debt or equity markets, you probably shouldn't be doing it anyhow. Three, I think it can make you lazy.
I think you can take comfort in, well, now we have all this cash, or you can spend it on R&D projects that don't really have the returns that you want to drive your overall ROE to. Like Pat was saying, ROIC is something we look hard at. My point is, I think there's a philosophical divide amongst certain management and certain investors as to what is that signal. For me, it's not a signal that there isn't something better to do with the cash. It's a signal that what we're doing with our cash is so darn good that we actually can give some of it back to the people that let us use their cash to build the company. I think it's just a completely different viewpoint on things.
Yeah, I'll just echo one of the things Steve said, that when he talked about the next five years may be like the last five. Because we all said the last five weren't great, even that last five was near $10 billion. That was the pre-consolidation.
No, the last five, I didn't say they weren't great. I just said they were hard. It's hard.
Yeah.
It's really hard. Look, it was great.
For R&D, it is. Yeah.
it's a big company, but it's a hard environment to manage in.
All right.
Okay.
We'll take the Q&A.
Q&A.
Thanks very much. Following right off that last thought, I guess, strategically, if you go back, like, I don't know, at least a decade, if not 15 years or so, old Seagate, there was a time when you guys used cash and actually were investing in, I'll call it non-drive technology. There was like-
All that software crap we bought-
You had Zoox and God knows, I don't know, probably 50 things.
Well, Veritas wasn't a bad one, but go ahead.
Right. No, no. I don't mean that as a good or a bad. I just mean that was obviously one strategy, which was take all this and invest in other storage sort of more broadly, and it would seem now that not that.
Not really.
I'm just wondering what the thought process was.
Well, let me recast what really happened. That's an interpretation from the outside, and I can accept that because you weren't in the boardrooms and with Al when we decided what we were doing. What we were doing is, look, one of the things that Al never gets credit for because he's so closely associated with being a disk drive engineer, is Al saw the implications of client server. He really had a fundamental understanding of what was going to happen to computing needs as the result of client server deployment. He was thinking about it from the perspective of an individual that clearly was a driving force in mainframe computing.
He had this worldview about, wow, once you start deploying technology in more unit scale and cheaper and enabling more people to access technology, oh, by the way, isn't that a lot like what's happening now with smartphones and iPads? That created a massive opportunity for technology companies around hardware and around software, and the integration of the two. It wasn't about getting away from the drive business, it was understanding that deployment was going to create a massive opportunity for the drive business, including the software implications. Therefore, let's participate in that because the quicker we could help that advance, the quicker that deployment was going to happen. That's what drove us. Originally, our whole software strategy, which then I was hired to lead, was around information management, storage management, network management software that was designed around the client-server architecture.
In that sense, if you say, how do I overlay that with what I talked about today? Yeah, we're going to do the same things, but it's a different animal. It's around mobile and cloud. It's around consumer and commercial convergence. It's around open-source software changing the entire underlying device technology opportunity. Yes, you can expect us to make investments, whether or not that's partnerships, whether or not that's JVs, whether or not that may be technology acquisitions for some software stack that we think we might need. We'll do that. Look, in terms of that report card, we invested $400 million over seven years, and that created $22 billion of value. I'd be pleased if we could do that again on the same scale. We're not going to shy away from it. We think it's an opportunity. We'll be prudent.
We think we're pretty good at it. I think it's an opportunity for us. I think that's what's great about being in technology for as long as I've been in it. When I started, I was doing Fortran code for mainframes that only 2% of the population had access to in a business. The world's completely different, and that's good for all of us. We want to be aware of those opportunities, but always leveraging off of our core strength, which is storage. Al did have that thesis as well.
Keith?
Whoever gets the mic.
Hi. I guess I'm the lucky winner. I had two. First, one of the premises for your model is no share loss. Toshiba's been pretty active about talking about they want to gain share and are willing to invest even in their partners to help their supply chain. How do you reconcile their statements with both WD and Seagate saying their share stays where it is?
Look, people forget a lot about the drive industry. First of all, I think people are very aware that it's funny. People think about it individually right, but not necessarily altogether right. There's a bunch of people that understand how much technology this requires and what that means in terms of fundamental ownership and development, advancing technology. Unless you're doing it today at volume, how do you do it tomorrow at volume? That takes a lot of expertise. Vertical integration has its advantages, especially as we move forward because the interrelationship between heads and disk is becoming more and more important. The material science around how a head works and how that interacts with the disk and the magnetic layering on the disk is becoming very highly correlated. Having that capability directed by internal resources is a big advantage.
It doesn't mean that you can't do it with partners. It's just a little bit harder. There's the scale of the investment that goes to support the R&D. There's a lot of capital that goes into supporting the R&D, and therefore you have to have a business model that says, "I can make that investment." Again, an entity that's getting the return on the total value of that probably has a better ability to make that deployment than individual companies, which is why there was first consolidation upstream, now here. The second thing is that we're a high-volume manufacturing company, and I think that's, again, a lot of people maybe sometimes focus on that in terms of the, what's the TAM and absorption? It's that combination that's really the magic of it, right?
You're the only industry in the world that is as high tech as we are across as many scientific disciplines as we are, at the volumes that we are. We make 2 million disks a day. It's like these crazy numbers, right? If you're not at scale, if you're half as big as the other two people, you aren't at scale. It's really, really hard to do that cost-effectively. If you're at a parent company that's generating, let's say Google owned them and said, "I don't mind losing money to grow scale." They're not owned by Google. They're owned by a company that is challenged on every one of its main strategic fronts in a substantial way. You have to say, is that where they're going to deploy their precious capital versus some of their other opportunities like fabs? My bet is no.
If I'm Toshiba, and I'm thinking about I have to beat Samsung and whoever else you want to throw in the list of fab guys, my focus is that's my business, right? That's where my dollars are going to go, not trying to scale my disk drive business to compete against guys who are twice as big as me with deep technology cores. Do I think that they can be successful being a player in the 15%-20% range by focusing on a limited offering in the enterprise and a limited offering in notebook or hybrid? Hybrid would be a natural spot, obviously, for Toshiba to be competitive. Absolutely. I think that's what we assume will happen because that's going to be the highest return for them. Trying to buy market share, for what reason doesn't really make sense because the margin model's not going to support it.
We'll see. I don't run that business, so we'll see.
Point out that I ask my questions one at a time instead of all at the same time.
Yeah, it's no problem.
The second question is.
Makes it easier for my fading memory.
That's why I'm working with it. The second question was on areal density. You've talked a little about, and you've talked about a 5% kind of revenue growth over time. Could you bridge the gap a little bit, at least to help us think about how you're approaching the unit volume growth for both yourself and the industry to kind of pull those two together?
Here's the good news. We don't have to think about a long-term unit growth right now because we're not deploying capital based on what we think is going to happen in three years. If we were leaning into capital deployment, we would, right? That would be a big point of discussion. Is it going to be 600 or 700 or 800 or when's it going to get to a billion and when are we I think that's the importance of this issue that Sherri raised and one of our investors raised about, we're going to lag. The cost of lagging is you may leave revenue on the table. Right? Because if you lag and the demand is better, then you say, "Man, I wish I had those extra 10 million units to sell this quarter," right?
Especially if the demand was 15 more because it wouldn't be even screwed up pricing. That's just not where the industry is right now in this macro environment. I don't really need to make a big, long prediction about it. I need to think more about petabytes because I have to make heads and disk, and I have to invest in heads and disk. Our drive asset footprint is the easiest one to manage both up and down. It's the least expensive, and it's the easiest to manage.
I have to be more focused on my short-term forecast about that. I don't want to whipsaw Dave and start a quarter at 70 and then tell him, "Well, no, I mean it's 48. Oh, actually it's 63, and then it's 52 and a half." I mean, that's terrible for him because he's bringing people to work, then he's sending them home, and he's bringing them back. There's a lot of issues around our day-to-day quarterly management. I mean, our master schedules change every day. I mean, every week, but every day. I think right now we look at it as, again, marginal growth in the December quarter, probably in that because I think September is probably a bit depressed by inventory adjustments and things like that.
Whether or not it's 150 or 155, we'll see obviously Windows gets released, and we'll see what kind of impact that has. I think you probably stick with a pretty flat forecast for a while until you see if there's any dynamic that easing does or any other Chinese government transitions, and then maybe they start a stimulus program. I mean, look, there's a lot of things to wait to be seen. Right now, I'd say we feel it's going to be pretty flat. Fiscal. Yeah, I think that's right. Yeah. I think that's right. Yeah. If it goes up, we'll chase it. We can chase. We're good at chasing. We'd rather chase. Dave would rather chase. Ops guys love chasing. They hate cutting. Yep.
Hey, thanks. Just a question about the long-term model. Was wondering if you could give us some sense for what your assumptions are around pricing, mix, and maybe some of the cost takeouts here at the middle of the gross margin model right now.
No. I mean, come on. It's like, how many SKUs do we have? Our business just isn't that simple to say here, and here. Actually, this is another thing Sherri, you hit on that's really an interesting question about how many gigabytes do people really need. The thing that's confusing about it all right now, because that thesis is out there, "You don't even need the 500 gigabytes." Why is average capacity per drive in the environment now where you can buy a single disk 500 really cheap, why is average capacity per drive still accelerating to the rate it is? One of the reasons is people don't realize what you store isn't necessarily what the system needs. The system does all sorts of little reading and writing and, what is it, a gigabyte an hour or whatever crazy thing it is.
There's a lot that gets going on with your disk drive that isn't about just storing your data. Average capacity per drive is accelerating. I mean, we'll be over 800 gigabytes per drive, I think, in the not-too-distant future. I think the assumptions are that that fundamental demand probably continues, but that the client devices go more towards single disk and two heads instead of multiple disk, and that the enterprise drives probably go from four disks to more like five or six or seven or eight, or maybe we do a half high again and squeeze more in. That assumption exists that there's going to be a heads and disk expansion, average heads and disk per drive goes up substantially from where we're at today.
That would be something if you don't like, you'd kind of say, "Okay, there's risk in the model." In terms of pricing, it says that pricing is managed to support what this industry needs, which we think is more like 27-32 points, not 20-24 points. That's again, back to if you need to deploy three times the capital for heads and disks to get to just that seven zettabytes, at some point, we have to have confidence that we should make those investments. If we get a couple of years of 30 points of gross margin with things being in balance, we probably start getting that confidence. You probably start deploying, and you therefore probably don't lose a lot of revenue because we were so short the supply. I think it talks to a more efficient supply chain for sure.
I mean, again, the real leading customers are out doing some great work with us. I mean, Dave can talk to this a little bit about things that make our operations more efficient. They make their operations more efficient. They reduce inventory in the system. They increase velocity. I mean, that's just great for all of us, and again, it allows them to match end-user demand a lot better than they do today, which is how do they solve that problem today? They carry a lot of inventory. Okay, that's not a great answer. I think mix SKUs pricing. All of it is baked into how we view the world. Maybe we bake it into the other way, which is, this is the model that we know is required to fund where we have to go at a minimum.
The industry is at a point now where it should manage itself that way. Can you get caught by sudden spikes one way or the other? A flood happens, there's a massive shortage. Yeah, of course, that's going to happen. Again, I think the big difference versus five years ago, you don't have as big a risk that if demand falls suddenly or consistently, that you have an overproduction, because we'll adjust production. We'll just say, "Fine, the TAM's 142, therefore we're only going to build this many units." That's an absorption hit on gross margin, but it's not a pricing hit on gross margin. That's the test that's going to hopefully it won't be in front of us. Hopefully, we get macro growth and the world doesn't get into this declining role scenario.
I think at 140, where it feels pretty again, 140 is December of 2008. Maybe the world doesn't feel great, but it doesn't feel like December of 2008 to me. Do you want to add anything?
A couple of things. Steve talked about making things easier with our customers. That helps logistics costs, which is a big deal right now, our side and their side, whether it's freight or inventory holding or configuration complexity reduction is another big effort there to grease their supply chains. I think on cost side for us, there's things like automation that we can do to address rising labor costs in various places that we're often doing. We haven't really pulled the handbrake on all capital. There's, for example, capital for automation projects that are going on to further that cost reduction. There are raw commodity costs that fluctuate. We all had the rare earth problem last year. It's abated somewhat, but it's not back to where it was two and a half years ago.
Those are things that we continue to watch, try to engineer out a better solution for, but I think in general, we'll still be able to address a declining cost world, and that helps some of the elasticity that might help drive demand out there in certain markets we have.
Okay. There's a question over there. Yeah. We'll have her get some mic, sorry. Yeah. Did you grab it from her like 20 minutes ago?
Just two quick questions. One, is another way to look at your margin profile there, Pat, that even in the current environment, let's say next five quarters, we're sitting there at $150-ish million, that you could go and be within that target gross margin, target OpEx range, everything else applies, we'll just collect it later as we need to?
Yeah, I think Steve highlighted the one issue is you have to be responsive to what the real demand is, you just line it up. Then you don't go price to go create demand. That's the fallacy. If you say back off, you back off, the industry acts that way because they should, because that's the right way to look at your resources, then you just back off demand, I mean, the supply, it'll stay there. If the industry doesn't back off on it, I guess you only have a few people to look at. If the industry backs off on it should stay in that range.
More a longer-term question. You talked about a target mix maybe of how much you'd want to offer in terms of hybrid drives, should we assume then that any of your pure SSD offerings would be part of that 20% three to five years out, what else is in that mix? Is that just-
Oh, yeah, sorry. No, I was talking. That's a good question. I just meant in terms of stuff that you would think of as spinning disk today, that product line. I think by 5 years from now, 80% of that product line has got a hybrid alternative or maybe only hybrid. Enterprise SSD would be on top of that. I view that as a different market, as you know. I view it as a complementary market. It's not a storage device. It's an accelerator. It's something that can munch massive amount of data very quickly, which means the massive amount of data has to be stored somewhere and moved in and out very quickly.
There's a lot of aspects of the performance of that device that have more to do with enterprise workloads than they do anything else, which is why Samsung came to us and said, "Hey, you guys know more about this than we do. What does that code do, and how do we manage it?" I still believe that the enterprise SSD market isn't going to be that interesting until probably 2015. I still think we're on target to be the leading provider of enterprise SSDs by that time with our technology investments. Doesn't mean we're not in the market today. We have enterprise SSDs today. We're going to ship the third generation later this year. We're excited about the product, look, it's just not a big market right now, and it's not a profitable market. I like bigger, more profitable markets.
On the client side, I think that's a different story. We do client SSDs today. I think that's another secret that we were going to talk about out there. Under the Seagate brand, we do client SSDs. We have a couple out there that are just running or whatever they're doing. I'm not a great marketing guy. Scott, do you want to give the pitch on the Okay.
They're in a case looking beautiful.
They're in a case looking beautiful. We do have client SSDs today, look, the real key to client SSD clearly is can you take flash that's used in devices that aren't compute-grade, i.e., memory sticks, and do something, because it's really cheap, and do something to make them compute-grade? This is really what our DensBits investment is all about. If you guys crack the code, that's going to be a big advantage to Seagate and our ability to play in the client SSD space in a highly competitive manner. It just depends how much interface value-add stack goes on top of that device? If it's a memory stick, obviously we're not going to compete with Samsung and Micron and whatever makes those things. There's a good opportunity for us, and we view it as a complementary market. Yep.
Joe here, from Citi. I wanted to revisit the long-term margin guidance. I guess one of the reasons why you previously had a 400 basis point range was to consider the severe price fluctuations during demand cycles. Given that prospect of better supply and pricing discipline following consolidation, would you consider a world where the fluctuation is less or maybe the range of gross margin could be narrower?
Joe, I don't think now. If the world were a more stable place, yeah, sure. What's volatility in the market right now? Betas are off the charts. Volatility is high. I think it'd be irresponsible to do that in the macro environment that we're in. Do we think of our business in a more narrow range? Do we try and Yeah, of course. I'm not going to sit here and say that it's going to always be in that range, or it's always going to be at the midpoint of that range. The reason we give a range is because it's a range depending on circumstances. The point is, it's a whole bunch higher than it used to be, and that's driven by the demand for storage, where the industry's ability to ship to is, and consolidation allows for better demand-supply alignment.
That'll result in a better margin structure. How much better? We'll see. That's a function of how well does the industry behave itself? What does Toshiba do? Is there macro growth? What if there's another recession? Come on. You tell me the answers to all those things, I'll tighten up the range for you. Yeah.
Jayson Noland with Baird. Thank you. I wanted to ask about the mix shift from mission-critical to business-critical high-cap nearline drives for the cloud community. What are your expectations, and can you drill a little deeper on the economics of a high-cap nearline drive sold to a cloud provider?
No different. I don't even know that there has to be a distinction between it going forward. If you really believe in cloud build-out in public and private, and their infrastructures that are being used for enterprise-level computing, whatever that means. That can mean a what was the thing I just read the other day that Berkeley ran a DNA sequencing thing on AWS, right? What is that? Is that an enterprise thing? Is that some lab thing? I don't know. It's all the same. That's my point. The infrastructures are basically colliding, and they're going to be leveraged across all these different applications. I think the delineation between what's a nearline drive and what's an enterprise drive is basically going to go away. To answer your question, the margin profile is essentially the same. If it uses more heads and disk, it's good for us.
Yeah, thanks. Aaron Rakers with Stifel.
Hey, you already asked a question.
I did.
First half. Yeah.
A lot of questions on gross margin. I want to actually tackle the OpEx side.
Okay.
You talk about ±$400 million. I'd love to understand, given the variations in the demand TAM, how much of that cost structures would you characterize as being variable in nature, and how quickly can you flex that up or down, depending on whatever demand presents itself? Any color would help.
You can flex it down really quickly once you make the decision.
Right.
It's just the question of whether or not you want to cut. You wouldn't do that unless you felt that there was some sustained pressure on your OpEx. Obviously, again, given our model, covering $400 million a quarter in OpEx is not a challenge. If the world's changed dramatically and it was a sustained change, the second we make that decision, it gets cut. Seagate is not unable or doesn't lack the tools to figure out how to cut OpEx from $400 million to $352 million or $336 million or whatever it needs to be. I don't think that needs to happen. Look, we always look for efficiencies.
Right.
We're doing a lot of things to drive better business processes. Dave mentioned logistics. Look, there's a lot of things that obviously we want to reduce our OpEx, and we try to every day. In terms of whatever that good efficient baseline is, can you cut it by 10%, 15% quickly? Yeah, you can cut it really quickly. The question is, but when you do that.
Sure
It's going to take you a little while to recover it if you decide all of a sudden you need it back again. In terms of flexing up, obviously, interesting question. It's probably hard too, just because.
Getting people
There aren't enough good people to hire in the disciplines that you need them. This is back to one of the earlier questions. As it relates to our R&D investment, it's a challenge. It's a challenge to find the engineering resources that we need to find on a global basis to keep doing what we're doing. That's something that Bob and I think a lot about. Dave thinks a lot about it because a lot of engineering goes into the process side as well. Flexing up, in some ways, might be a much harder challenge than flexing down.
Quick follow-up. The enterprise discussion on the SSD side, you talk about 2015. Your closest competitor last week talked about being a PCIe SSD vendor as well. I would love to hear your opinion on that market. Is that also a 2015-
I think it's an important interface. I think PCI is definitely an interface that a lot of people are using, and if you're going to be in the SSD market, you probably should have that product offering. Interfaces don't matter to us. People get all hung up on, "Oh my God, fiber channel is fast." We don't care. It's a chip. It's writing to some media. That's not what we do. If that's the interface people want, great, we'll write some code to make it a PCI. It's not a big deal. I do think PCI is going to be an interface that a lot of people use and want, and we'll have a product that addresses that. Apps will be more important. Yep.
Great. Thanks. Bill Shope, Goldman Sachs. You've obviously given us a lot of color on demand trends over the next few quarters, and the client being the greatest source of near-term pressure. How are you thinking about enterprise demand trends? Looking at this particular quarter right now, did you see incremental pressure on the enterprise, obviously cyclical pressure?
I never think about looking at a quarter and making some prediction about what that means. I think, in general, having been someone that's been way more embedded technically on the enterprise side than on the client side, I'll say what I said earlier. This change to cloud is bigger than client server in terms of the implications for anyone that's dealing in the enterprise environment, I mean that in a positive way for the drive industry, because you're lowering the cost of computing. They're not after us. The drive is really cheap. I saw something the other day, I think it was another AWS ad, they're really good at marketing. $0.01 a gigabyte per month. It's cheap. $0.01 a gigabyte per month, this sounds cheap to me. Penny's not much money, and a gigabyte sounds like a lot. That's $0.12 a gigabyte per year.
We deliver HDDs at $0.06 a gigabyte. You're renting something for 2X what you can buy it for. No, because you can't buy it at $0.06 a gigabyte. EMC can, HP can, but you can't. What happens between $0.06 and $1? Software and services. Are you kidding me? That's under attack. Maybe the $0.06 goes to $0.05, we hope so, because guess what? People use more of our devices. It's that $1 that's under attack, that will happen, by the way, EMC's doing it themselves, as is HP. Everybody gets the joke. The question is, who are the winners and losers? At the end of the day, that lowers the cost of computing, which is good for us.
I think the more you bring that cost down to the real hardware and the true value of the software, guess what? Deployment goes up, I think this is bigger than client server. We're pretty excited about it on the enterprise side. My answer is, I think for whatever good that happens out of the client side with NAS devices and all that stuff and thin and lights and hybrid drives and five millimeters, I think what happens on the enterprise side is bigger.
With that in mind, if we look at the enterprise over the next several years, you obviously have one key competitor there you're focused on. How do you think about the sources of competitive advantage and what Seagate can do to have a longer-term advantage?
Look, they're a great competitor. IBM, really who that is a really good disk drive company. IBM and Seagate were the two best disk drive companies in enterprise, Hitachi continued that heritage. Whether or not WD does or not will be interesting to see. Steve now running the company, maybe it makes it more likely versus less in terms of the operating philosophy. Seagate's broader, has been broader. We do more form factors. We do more spin speeds. We do more interfaces.
More SKUs.
We have way more SKUs to support our customers, they're really, really good at what they do, that makes us better at what we do. This world is plenty big enough for BMW and Mercedes. Audi now. Man, that new Audi A8's really nice. I don't worry about them. I want them to be competitive. People asked me before that merger went through, did I hope it'd go through or not? You could tell that the answer they were looking for was implying that I was going to say whatever was the weakest result was what I wanted. It isn't. I wanted it to be the strongest result. We had a great competitor in WD on client that made us do things every day smarter on client. How Dave runs his factories, what's our product roadmap? What do they do?
Same thing with Hitachi and Enterprise. They made us do things better every day because of how good they were in enterprise. If by putting those two companies together, they get stronger, that's really good for Seagate because it's like any of us who play sports. You want to play the person who's really, really good, not the person that you can beat every time, because they bring out the best in you. If you don't believe that, well, none of you would be in the business you're in if you don't believe that. Your business is all about competition. My yield's better than yours. I think it's going to be interesting to see if that's a stronger company together than apart.
I think that under the regulatory rules that they're operating under for the next two years, that's going to be a challenge, because they can't get the leverage that they really want, whether or not it's technically or operationally. We're not going to presume that's the way it's going to stay. We maybe view that as a window that therefore we can even get more advantage so that when they do get together, we're that much more competitive. Look, they're going to be a good competitor going forward, and we're prepared for it. I think it's going to be a good head-to-head competition. Yep.
How do the buyback and share count targets change with changing stock price? Is the plan to get down to those share count levels irrespective of where the stock price is?
I think of it more in terms of multiples. If enterprise value to cash flow multiples stays super low, you just keep buying the stock. If they go high, maybe you slow them down. I'm not really worried about the equity market running away from us in the next two years. My guess is that we probably just keep buying. That's a fair question. Yeah. We're at like four times today? Look, I think if you're at six to eight times, that's I always think of it in terms of financing, right? If you can finance debt at whatever rate, maybe we'll be like Switzerland pretty soon. We'll be able to finance at -2%. Wouldn't that be cool?
Look, if you can finance reasonably at 5% or 6% or whatever that Seagate's rate would be today, people are looking at coverage ratios of eight times, I think that's where your breakeven is, right? If you're starting to get valued more than that means you can't really finance debt with that, which is how I think of it from a balance sheet perspective, something like that, 2x where we're at today. Like I said, I think we got the runway. Yeah. In terms of dividends and buyback?
Yeah.
Yeah. Yeah. Yeah, we did. We know your opinion. You wrote about it. No. Yeah. Yeah, I think as you saw on Pat's chart, you see the transition to that. The ratio between last year's to this year's is definitely shifting towards dividends. What Pat didn't show was a long-term philosophy, whatever that means. No, I mean, long term, I think the philosophy is Well, no, he said minimum 10% growth. He also said, and witnessed that we've done step function changes between now and then. You guys always try and boil it down to just one thing. That isn't what we said. He said minimum 10% as a model, and don't forget, we do step function things. Let me answer your question.
I think that where the board is at, and by the way, this is the board's job with all of your input, and there's a lot of diversity in input about whether or not you do buybacks or dividends or management bonuses. Although that's a really small percentage of the shareholders that are voting for that. I think where we're going to end up is probably around a 40%-50% payout rate with 80% of that dedicated to dividends. Okay? No, it doesn't look like we're doing it today. Why? Because of your thing. Because at three times cash flow, are you kidding me? Because if I take those shares out, I get there faster, right? Because I don't even have to increase the payout ratio, and I can get $2 a share, right, if I'm down to 250 million shares.
I think it's just where we're at right now. You and I do not fundamentally disagree about this, how you return that capital the best way, when the people capitalize what rate, blah, blah. Right now, because the disconnect is so big on valuation versus cash flow, I'd rather reduce the equity base so I can raise the dividend rate without increasing. Yeah, exactly. Well, exactly. Which is what we did last year, right? We had basically 100% payout last year. I don't think we're that far off. Okay. In the back.
I just have two quick questions. On the tax rate, what should we be assuming for a tax rate over the next few years?
Tax rate, I'd tend to look at it more as a fixed. If you're looking at the model, I'd put anywhere $6 million-$18 million in your model. $20 million a quarter.
Yeah, $20 million a quarter.
Perfect.
My next question regarding increase areal density slowing down and more heads and platters increasing. Typically, the largest driver of gross margin for you over time has been the ability to take out components. Now it sounds like w e're going to be entering a stage with increased component costs. Is that a significant gross margin challenge, especially on the enterprise side? How should we be thinking about that?
I think it's back to this question of, unless you have the gross margins that tell you to deploy that capital so you can maintain those margins, you're not going to do it. That's what our customers need to start understanding, right? That the reason that this new model is what it is, and that you're not going to run back to whatever it was when there was a different areal density to petabyte growth relationship and a different industry structure, is because that's what you need in order to make those heads and disk. Yeah. The component cost will come down. Oh, yeah. Component cost comes down in all those models substantially. You're adding more heads and disk to get to the Again, we have customers saying, "Give me 20 terabytes under one spindle." No problem. I can manage it now.
Three years ago, people were saying, "I can't manage more than four terabytes under a spindle." There's been a huge breakthrough in terms of the software capability. Long-term, that's a really good thing for Seagate. That's absorbing heads and disk capital, which is what we want to absorb. I tell you where it gets tricky. Dave alluded to this. The first place it's going to get tricky is media, because the first shortage isn't going to be drives. The first shortage is going to be media. Media plants are expensive, and they got long lead times. How far into the shortage are you before you say, "Okay, I'm going to build another plant"? That's going to be an interesting question. The second big break point is probably actually slider before wafer even. Slider. The slider, and then wafer.
Those are decisions that have, literally they have at least a year of lead time if you have the green field picked out. When the shortage comes, I say it's going to make the flood look like a hiccup because it's not something that's going to get resolved by rebuilding your supply chain really quickly in six months. These are long lead time capital deployments if you have a building ready. What are we doing? We're thinking about ways that don't require buildings to be ready, and we can do it more incrementally. Okay. Questions? Yep.
Yeah. Nehal Chokshi, Technology Insights. Western Digital announced last week that they will be coming out with helium-filled drive.
What?
Can you talk about when you guys expect to do that?
I was dying all day for someone to ask that question.
Well, can you talk about when you expect
It's a radically new technology. We were blown away. Bob's scrambling to see what we can put together. Wait, sorry. I'm supposed to be polite. Okay. We've been building helium drives for 20 years. We've got most of the early patents on it. We use helium every day in our production. Given the technologies that we have to deploy, we don't see that as a necessary technology, given the risks that are involved with it as what's required to meet what customers are asking for, but we certainly can do it. Bob has a funny story about it.
Yeah, I probably shouldn't go down that path.
No, go ahead. It's a good story.
Well, actually, the biggest thing is the cost. You've got to get all those strings for Dave to tie, so they don't float to the ceiling. No, seriously. I think Steve's point is just that we've had a lot of experience with helium. We certainly know the benefits. There's increased costs for doing it. There's technology changes that you have to make, and at this point, it's just not something that we need to do. If we wanted to, we could definitely deploy it down the road.
Okay.
Yep.
The follow-up to that is, so the 3.5-inch technology, the runout problem, that's not going to be an issue without needing to have them permanently helium-filled. Is that correct from your view?
We don't see that.
Right. If you look at it, they announced a 7-disc, 1-inch drive, and you can do air technology, current technology, without using helium on 6 discs. It's about a 16% capacity improvement. There's lots of ways to go solve that equation. For example, I mentioned earlier that we're going to ship Shingled Magnetic Recording this year. That's over 16% benefit with just what I'll call, and it's not just but firmware, without changing the whole chemistry of the drive, without having to go to radical seals, without having to go to super thin discs. There's other ways to get at 16% capacity. If we wanted to down the road, we could go implement that.
Yep. Okay. One in the back. Yep.
Kulbinder Garcha, Credit Suisse. You quickly touched on your M&A strategy. As you look forward to entering the enterprise SSD space by 2015, how do you view build, buy, partner?
Sorry, I think I got it, was with respect to SSD, how do we view partner, buy, or make for SSD? I think it's the relationship between the media provider and then the drive provider on the enterprise, our knowledge of workloads, interfaces, test, that provides a lot of value. You want to be able to leverage whatever controller technology that you have to go as deep into the silicon as possible. The deeper look that you have into the cell structure gives you the best ability to have those devices not burn out, and that's what our partnership with Samsung is about. I don't see there being any dramatic change to our partnership with Samsung.
We picked Samsung because we believe they're the best long-term player in developing that silicon technology, and they were willing to basically show us their secret sauce of what's going on at the cell level. I don't think there's going to be a big change in that. Can you see technologies that you may want to buy or invest in? DensBits is a great example of that. That's a technology that, one, we felt that they were pretty far down the path of solving some interesting technical issues that if they're successful in, will make a big difference in what kind of silicon you can use, and we wanted to have rights and ownership to that technology.
Yeah, I would say there's really three factors in how we look at the SSD market. The first one being the sensible technology, the second one is getting time to market, and the third one, getting time to scale. We look at all three of those factors in figuring out how we're attacking the sector, and I think our decisions all support that kind of priority stack.
Yeah. I don't foresee any big changes in that strategy. It'll probably mostly be partnerships, maybe some investments, and then continue to develop the program that we have that we feel pretty good about. Yep.
Two questions. One in terms of how to think about market share over the coming year. Seagate experimented with auctions this past year. If you could update us on how you're thinking about that and how the LTAs, as they expire, will play into market share. The second one on SSDs, if you could look at strengths and weaknesses between, say, a flash provider and at Toshiba, which is really coming from both the solid-state hard drive side, and a Western Digital or a Seagate coming with the hard drive side in terms of who's got what advantages.
Well, like I said, if I were running Toshiba, you'd think that between leveraging the Fujitsu enterprise capability and the Toshiba wafer capabilities, that you could have a pretty compelling offering in enterprise SSD. I think that'd be a good area for them to focus on. You have to decide how they're going to do competitively against Samsung and the fab world, their lineup looks pretty good. Their linear density roadmap is pretty compelling. I think in terms of WD and Seagate, well, Hitachi more than WD. If you meant the old WD, then that would be a really easy answer. Hitachi, because of the Intel partnership, I think clearly has a very good product today. They have good traction, they have good engagement with customers, I'm sure they'll leverage that. For us, it's Seagate and Samsung.
We decided to say the leading enterprise player should be with the leading silicon player, and together we should be pretty successful. I like our odds. Like I said, I think by 2015, 2016, we'll be the leading supplier in that space. It's going to be a competitive marketplace. There's some more questions there, I think. No, you. Yeah. Again, I talked about the Long-Term Agreements. The Long-Term Agreements, even though you all wanted to think of them as these share lockup things, they were more about, again, engaging with each company's supply chain to make both of us more efficient together. Regardless of whether or not there's some contract with words, the relationship is formed. I don't see any big shift in market share because LTAs do or don't expire. By the way, we have LTAs that go for two years and three years.
We also have people that just re-up their LTAs for one and two years. They're not these things that are defined as I think investors want to believe they were. They're really good contracts for us, and they were really good contracts for our customers, and I think that's what people don't understand, that smart customers realized that the ability to have access to this technology became very important to them when there was a sudden supply disruption. They don't want to get caught in that trap again. Embedded in our belief that we can hold 42% market share, plus or minus, is our knowledge of what our agreements are with our major customers, regardless of whether or not that's documented as an LTA or not. Yeah, that's included in that thinking. Auctions, I don't know if we're doing any more auctions.
Rocky should talk about auctions the most. To us, it was another test of how we get product to market, and I think we're going to continue to develop the concept. We may use some different vehicles to do it. Yeah, we're still experimenting with it.
Yeah. Exactly what Steve said. We think it's an interesting channel for sales, and we're going to continue to research and develop it. Some quarters we're going to do things that are meaningful impact, and some quarters it'll be a learning experience, but we're pretty committed to continuing to develop the platform.
Time for what, one more? If there is one more.
I have a different question.
We don't have any planning horizon that would say that we would do that. When we look going to dividend, we're saying it's a fixed charge. I guess if we couldn't fund our product portfolio or deploy capital as we need it, then you'd have to look at that. It's right after that, and with us supporting our debt. I don't see a scenario in our planning horizon that we would cut that.
Yeah. When we established the dividend, we're definitely establishing it with the mindset that it's there or better going forward.
Which is why we stay away from the one-time dividend thing. I think to me, there's a lot of interesting debates about what happens if you win the WD thing, and all of a sudden there's $700 million that you didn't have yesterday. Do you dividend that out or not? I think we probably just say this is just more capital that we could then sustain the dividend growth that we don't ever have to backtrack on. That's for the board to decide. The philosophy is when we set our dividends, we're setting it to be there or better.
Based on everything that we know. If the world cratered, of course, if you're saving the company, then you save the company first. All right. I want to thank everyone. You spent a lot of time here. I know a lot of you came from as far as London, very impressive, and I really appreciate it. We appreciate your support, and we look forward to developing our relationship further. Thank you.
Give them a key?
Oh, everybody gets What's my pitch?
iamaKey.
iamaKey.
You give the pitch. iamaKey?
Yes. As you walk out, please look for the people that are handing out one of our latest key products called iamaKey.
Thanks, everyone.
Thank you.