Good morning. My name is Karen, and I will be your conference facilitator today. At this time, I would like to welcome everyone to the Micron Technology SSD Business Update conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question-and-answer period. If you would like to ask a question during this time, please press star then the number 1 on your telephone keypad. If you would like to withdraw your question, please press the pound key. Thank you. It is now my pleasure to turn the floor over to your host, Micron Investor Relations Director, Ivan Donaldson. Sir, you may begin the conference.
Thank you. Welcome to the Micron Technology SSD Business Update. On the call today is Glen Hawk, Vice President of Micron's NAND Solutions Group, Justin Sykes, General Manager of Client SSDs, and Ed Doller, Vice President and General Manager of Enterprise SSDs. This conference call, including audio and slides, is also available on Micron's website at micron.com. Our call will be approximately 60 minutes in length. There will be an audio replay of this call accessed by dialing 404-537-3406 with a confirmation code of 93147697. This replay will be run through Wednesday, August 15th, 2012, at 5:30 P.M. Mountain Time. A webcast replay will also be available on the company's website until August 2013. We encourage you to monitor our website at micron.com throughout the quarter for the most current information on the company, including information on the various financial conferences that we will be attending.
Please note the following safe harbor statement.
During the course of this meeting, we may make projections or other forward-looking statements regarding future events or the future financial performance of the company and the industry. We wish to caution you that such statements are predictions and that actual events or results may differ materially. We refer you to the documents the company files on a consolidated basis from time to time with the Securities and Exchange Commission, specifically the company's most recent Form 10-K and Form 10-Q. These documents contain and identify important factors that could cause the actual results for the company, on a consolidated basis, to differ materially from those contained in our projections or forward-looking statements. These certain factors can be found in the investor relations section of Micron's website. Although we believe that the expectations reflected in the forward-looking statements are reasonable, we cannot guarantee future results, levels of activity, performance, or achievements.
We are under no duty to update any of the forward-looking statements after the date of the presentation to conform these statements to actual results.
I'll now turn the call over to Glen Hawk.
Okay, thanks, Ivan. Hey, to set the stage, I'd like to thank everybody for joining us on the call today and also talk a little bit about the growth opportunity that Micron sees in the NAND industry, and in particular, why we're focused on SSDs. If we go ahead and go to the first slide. We like to look at the market growth in terms of gigabytes. This is important to us because it's directly proportional to the amount of capacity we have to put in place, the wafers that we have to run in our multi-billion dollar factories. As a figure of merit or reference by comparison, our highest density NAND chip today is a 16-gigabyte chip, and we measure the total industry in billions of those gigabytes.
As you can see here on the left-hand axis, we're showing demand total market growing from somewhere in the 30-ish billion gigabytes today to something around 160 as you look to the right in 2016. Just phenomenal growth. Now, I don't want to let this exponential curve alarm any of you. For our industry, this is actually quite normal. In our industry, the name of the game is to lead with NAND technology, get to the next lithography as fast as you can. Overall for the industry, we're able to do that about every two years. That means that we can roughly double the number of gigabytes that we can produce on a fixed capacity base every two years.
If you do some math on this overall curve, you'll see that we're projecting something well north of that doubling every two years, which would be about a 40% compounded annual growth rate. The curve that we're showing here overall is closer to 50%, in fact, a lot of our peers in the industry, the analysts, et cetera, forecast things sometimes closer to 60% or even higher. The point overall is that the NAND market's growing, in addition to racing to the next lithography, eventually the industry's going to have to put some additional capacity in place. I'd like to call your attention to the three colored regions on this graph.
What you see here at the bottom in the gray regions is sort of the traditional or maybe the original growth drivers for the NAND market, this was removable storage, SD cards, USB devices, things of that sort. As you can see, the bit growth for those segments does continue as we move out in time. You see some other segments emerging, however. You also see that later in time, mobile became a large segment that consumed a lot of NAND bits. One of the ones that you actually see emerging there around 2011, 2012, is the dark green region, which represents tablet market. What we're here to talk about today are the blue regions here shown at the top. As you can see, these are the most recent emergence.
The CSSD and ESSD designators here are designed to indicate the growth that we see in client SSD, Ultrabooks, et cetera, the Enterprise SSD category represents the SSDs that we see in the even more recently emerging opportunities in the data center. What I'm showing here again is the gigabyte opportunity. Later on in the call, Justin and Ed are going to talk a little bit more about the dollars. The key point here is that this most recent emerging segment here, SSDs, it's very exciting for us at Micron. If you go to the next slide, the reason that we're very excited in this opportunity is twofold. One, solid-state drives represent a more fully integrated solution. That gives us, as a NAND manufacturer, the opportunity to participate in additional value that our products bring to the world.
Secondly is the emergence of these new segments. It corresponds with a point in time that's very challenging for the NAND industry in terms of scaling. A lot of you may have heard that the entire industry is having an increasingly difficult time scaling to the next generation and keeping that exponential increase of bits growing for the world. What that means is that in order to deliver solutions going forward, those who have the most intimate knowledge of the NAND technology are going to have an advantage in terms of bringing leading-edge products and solutions to the world. As a result, a big focus for us at Micron is capitalizing on this window of time where we have this opportunity to climb the value chain.
What we have been doing is we have been building upon our base of leadership in silicon technology and NAND product design and development. We've added on top of that the ability to do our own controllers, both internally as well as externally with partners that allow us to manage that NAND media. That's where a lot of the IP and the secret sauce, so to speak, that's required to deal with the challenges that I mentioned earlier with the NAND technology really come into play. On top of that, we've been able to add the other elements that are required to bring total system solutions to market in the form of our SSDs, again, our client and our enterprise SSDs, which we will talk about.
We will also touch on at the end an even higher step in the value chain that we're entering, which is the actual delivery of our own integrated software and hardware appliances, and Ed will touch on that at the end. If we go to the next slide, I'd like to just touch briefly on each of these steps, so to speak. Most recently, you may have noticed we've had some very nice proof points for our technology leadership. Earlier in the year, we were awarded the prestigious UBM TechInsights Semiconductor of the Year award. Most recently, we also made a big splash with the introduction of our first-to-market phase-change memory solution for the mobile handset market.
If we go to the next slide, we also within the last year recently introduced our leading-edge 16 gigabyte NAND device, the one I mentioned earlier, a nice photo of that shown here, and overall, we have a complete portfolio of a broad range of densities, and it includes Single-Level Cell, Multi-Level Cell, and Triple-Level Cell products. We also deliver our NAND components in more highly integrated form factors that require controllers packaged within the solution. For example, embedded MMC-managed NAND solutions. If we go to the next slide, I also wanted to reiterate that with regards to the controllers, we have been shipping our NAND products with controllers for quite some time now, and as the challenges have increased, so has our internal capability on key core competencies like error correction and digital signal processing.
We're now in the position where a lot of our products that are shipping today are shipping with our internal controllers as well as those that we work with our partners on. If you go to the next slide, overall, this has enabled us to enter the SSD market. Today, we have about 10% of the worldwide SSDs for client and enterprise combined. Again, Justin and Ed will speak some more about each of those in the next section. Finally, here is a nice photo of our integrated software and hardware appliance that I mentioned earlier. This was brought about by an acquisition that we announced earlier in the year of Virtensys. This product adds virtualized IO appliances to our product portfolio, and it enables data centers to share local storage across multiple servers. Again, Ed will touch on that some more.
With that's a very exciting time for us here at Micron. What this means to us from a revenue or revenue quality perspective is that as we move to the right, you're going to see an increasing percentage of Micron's NAND revenue go toward SSDs and these more fully integrated devices. Essentially what we're showing is that between the years 2010 and 2014, you can expect to see that percentage increase from about zero up to just over 40% over the next few years. That, of course, will have a very positive effect on our bottom line as well. With that, I will go ahead and turn the floor over to Justin.
Thanks, Glen. Next, we'll spend some time and drill into some more detail on what we have going in the client SSD market. Go to the next slide. First off, we're seeing some trends and shifts in the market that are really quite exciting for us with the introduction of the ultrathin or Ultrabook devices in the notebook space, as well as the continued growth of tablet PCs in the market. What we're seeing is by 2015, we expect that this market is going to grow to be about 60% of the overall client PC system shipped in the market. The thing that's really key about this is both Ultrabooks and tablets require some form of NAND for their storage device, whether that be the primary storage or through caching.
This is really exciting to Micron and drives tremendous growth in the overall NAND market, as you saw previously from Glen. Looking at that in a little more detail, on the tablet space, we expect to see tablets grow on the order of about 76% between now and 2015. Those will be raw NAND. Some of it is the managed NAND solutions that Glen talked about, and a portion of it is also SATA SSDs that will be consumed in these devices. Finally, with the Ultrabook introduction and thinner and sleeker notebooks, highly appealing to consumers and on a very fast growth rate this year and predominantly going into 2013. These are using SSDs for primary storage, mostly through the new module form factors called mSATA, as well as using NAND in a cache, either as an SSD cache or in a hybrid hard drive.
Go ahead and go on to the next slide. Looking at this in the overall market, and this chart, what you're seeing is our view of the overall SSD market in terms of revenue dollars. If you start at the bottom and look at the two blue pieces, really this represents the notebooks and tablets that are using SSDs in primary storage. The notebook piece is predominantly driven by Ultrabooks, although we do see a lot of SSD use in traditional notebooks as well as a hard drive replacement in an upsell. The tablet piece is really just the portion of the tablet market that's using a SATA SSD. We're not looking here at the eMMC or the raw NAND consumption, which is also quite large. Moving up, actually, to continue on the notebook piece, we really see this as the larger growth in the market.
It's predominantly driven by the OEMs and new system sales. As more and more of those systems transition to the Ultrabooks, we definitely see continued revenue growth in this space. Moving up to the green section, this is the channel market. We define the channel market as e-tail, retail, distribution, as well as some system integrator-type customers. This is a market that Micron has been quite active in. We see it as a large portion of the overall SSD market and continuing to be a significant portion going forward, although not growing at the rate of the notebook space. Micron predominantly addresses this market through our Crucial brand. For those of you who are not aware, Micron has a brand through the Crucial that is more focused on the e-tail market.
Really, if you go out on the websites and look for SSDs, that's really where you find the Micron product is through that brand. That's been quite successful for us, and we continue to play in that channel space with that brand going forward. Moving up the chart to the next rung, you see the cache market. There's been a lot of debate in the cache market on whether the caching will be hybrid hard drive or dual drives with an SSD. For argument's sake, what we've done here is just included our view of the total cache market, whether it be a hybrid hard drive or an SSD. It is a growth market. As Ultrabooks come into the market, they need caching or primary storage.
We see a lot of unit volume here, but the gigabytes in the caches are quite small, 24, maybe 32 gigabytes, compared to, on average, 160, 170 gigabytes for primary storage. As a result, the cache market, while large and driving a lot of unit volume, is actually relatively small in terms of revenue, and we see it kind of peaking out in 2013 and 2014, then it really starts to slightly decline as we see primary storage eating into that caching market. Finally, there's a small section of other market, which is really kind of industrial, more the historical, traditional SSDs in the market.
At a top level, we do continue to see fairly strong growth in terms of revenue and especially in terms of units in the overall market growing from this year at about $5 billion to just around $8 billion by 2016. Pretty exciting growth that we're seeing in this market. Go ahead and go to the next slide. Drilling into what Micron's doing in this market. Historically, Micron entered the SSD market in a real way, really in about 2010. We entered that market in the channel space through our Crucial brand. If we play this back a year, it'd be 100% channel for Micron and Crucial brand for Micron in 2010. Through that year, we began to engage the OEMs in the space and were able to successfully get some qualifications.
By 2011, we were shipping into some OEMs, and that began to become a more significant portion of our shipments. As I showed you in the growth chart, that's really where a lot of the growth is coming from in the market going forward, and it's definitely a big focus for Micron going forward. We will project to continue to grow our share at the OEM space, and as a result, it's going to continue to consume more and more of our output. By 2014, we see the OEM piece being about 60% of our output and the Crucial branded channel play being about 40% of the output. By no means is the channel going away. We continue to see that as very important to Micron. The OEM piece is also where we see the growth going forward.
Go ahead and go to the next slide. Some of the proof points around the channel market. Really, a lot of the press we get there is through the online websites, the review sites, who are looking at the products that are out and available in the channel market, predominantly through places like Newegg and Amazon. The C400- It's been very successful product for us there. We've been fortunate to get some very positive press reviews. Through the life of the product, even as recently as this year, even after it's been on the market for over a year, we continue to get high rankings from the press in terms of quality and performance, and continue to even win awards with the product as early as earlier this year.
We've also seen some good press out of some of the market analysts, really recognizing that Micron has taken a leadership position in the client SSD market, and continue to see that kind of press and hope to see it moving forward as we continue to release new products. Finally on the OEM space, as I showed you, we have had some success in the OEM market and continue to grow that success. We see that as very key to our success going forward. We're at a point now where we have been qualified in five of the top OEMs in the world, and continue to grow that presence and look forward to adding more customers to that list over the coming year. Go ahead, next slide. Kind of a high-level look here at what we're doing from a product standpoint to enable that success.
We've really narrowed our product focus in on main storage, and we see really three unique kind of segments in that market, starting from the entry and going on up to higher performance solutions. As you can see on the left-hand side of the chart, as you move up from the entry up, you really get higher and higher performance products. The flip side of that, you can see on the right-hand side of the chart is as you go up in performance, it also brings up the cost or the better way to look at it as you drive down into the entry portion of the market, you drive down cost. There's trade-offs and different portions of the market that can be addressed by each. Today where we are is in the gray boxes or shipping products.
We have our C400, which is also available through the Crucial brand as an M4, that really was a product that was brought out to address the mainstream of the market. That's really the only market that was out there early on is really this mainstream, or you might even consider it an enthusiast market. That's really, as over time, has come down to more mainstream acceptance in the market. As we've seen the growth in the client SSD market, there's really been a diversification of the needs of the market, and we continue to see that need in the SATA market for the higher performance product.
We also see a need for products that are more focused on cost, while still delivering all the excellent benefits of an SSD through lower power and relatively high performance and ruggedness and quality, really absolutely focused on getting some of the cost out of the system. That's been very successful for us. As we go forward, we continue to push down that strategy in the SATA space with two unique product lines. Moving into the next few months, we'll be releasing our 20-nanometer based products, again, focusing on performance for the mainstream and focusing on cost for the entry. Over time, we'll continue to drive cost reduction in that space. There's really been a desire in SSDs to continue to drive to higher and higher performance.
The SATA bus, we're at a point now where we saturated that bus and we can't get any more performance out of it. The next thing on the horizon that opens up that bottleneck is really PCIe, and we see an emerging market for PCIe in the very high-performance enthusiast portion of the market initially, over time, that'll waterfall down into the other segments. This is actually really a great story for Micron because as you'll see when Ed talks, we have a very strong play and position in PCIe and enterprise. We have a lot of knowledge base around the enterprise PCIe space. We can take that knowledge and leverage it down into the client space as we move forward.
Really feel that this portfolio puts Micron in a position where we're well-positioned to meet the needs of our customers now and well into the future. As we all know, the benefits of SSDs are great with the much higher performance, the ruggedness, the quality, the power. There's really nothing a person wouldn't want about an SSD in a system. It provides so many benefits. There's been one real inhibitor, and that's been the cost. The cost of SSDs has been quite high compared to hard disk drives. I think for this year, we've really seen some pretty exciting trends. We all knew the costs were going to come down, but what we've seen is primary storage, at least enough capacity to be primary storage where a person could use it in a notebook and really not feel like they didn't have enough storage locally with them.
Those drives have broken through this $100 barrier. The $100 barrier is significant to me for a couple of reasons. One, if you look at the channel market, there's this emotional attachment to $100 and this point where if you can pick up a product for less than $100, you really don't have to go through the whole thoughtful approval process. That purchase process just is that much easier for people. The other side of it in the OEM space is we've gotten to a point where the primary storage for a notebook can be had for roughly 2x what it costs to get a hard drive. That gap is continuing to decline.
As we play this forward and roll out continued cost reductions in the SSD space through process migration, as well as just bringing down the total cost of the unit, I can definitely see a point in the not-too-distant future where we'll be able to have this primary storage device, at least on a per-unit basis, on par with hard disk drives. That's going to be really tremendous for the industry and open up a huge growth in demand. Go ahead and go to the next slide. As you've seen, Micron is keenly focused on continuing to be a dominant player in the client SSD market through growing our portfolio of products. We're continuing to grow our customer base with a targeted approach to ensure we continue to provide our customers with the highest quality levels and support.
We're able to do this through our vertical integration approach. As a NAND supplier, we are uniquely positioned to leverage our knowledge that we have around the NAND silicon and build that out into SSD products through integrating our NAND knowledge into our SSDs, through firmware development, through the controller, through the SSD design itself, on into the qualification of the SSD and the manufacturing of the SSD. Really looking at it as a very holistic, from silicon to complete system product. This is something that is going to enable us to continue to provide very high-quality, feature-rich SSD solutions to our customers moving forward. With that, I'll turn it over to Ed to look at the enterprise.
Okay. Thank you, Justin. Good morning, everyone. I'm Ed Doller, Vice President and General Manager of Micron's Enterprise Storage Business. I'd like to take the next 10 minutes or so to discuss Micron's approach to the enterprise storage market. Please, next slide. If we start by taking a look at the enterprise SSD TAM, I like, obviously, to look at it by interface. We have it broken down by SATA, SAS, and PCIe. If you look at SATA over calendar year 2012 to 2016, you see that it has about a -13% CAGR over this period, while SAS and PCIe are showing CAGRs in the mid-30s. I believe SAS is at 36% and PCIe is at 33% over the same period of time. Both SAS and SATA are interfaces that certainly have been around quite a while in the interface.
The result of this growth, in my opinion, is twofold. Number one, it's because the total cost of ownership model in the industry is being much better understood by our customers and data center customers as well. Of course, I think as NAND $ per gigabyte cost decline, that total cost of ownership certainly becomes more attractive to more customers. PCIe specifically, is certainly a much newer and more exciting interface and gives the server OEMs a new view of storage. I'll likely spend a couple of more minutes discussing this interface as I walk you through Micron's approach to the enterprise storage market. Next slide, please. As I noted on my previous slide, PCIe is new and most importantly, puts storage back into the server.
For those of you that recall, it wasn't that long ago then that the world moved DAS or direct-attached storage out of the server given its growth. In fact, I think I was actually at IBM when this transition actually happened. I think this new movement is certainly not going to move storage back in, but it's going to challenge the conventional storage OEMs for value, hence my storage wars title. If you take a step back and look at this from the high level, I think what we'll start to see is higher densities and higher performance PCIe drives making their way into servers, first under the guise of cache, and the value of performance in the NAS or SANs will certainly be challenged. I'm not saying that they will necessarily compete for gigabytes, but they certainly will compete for value.
For example, if your cache becomes big enough, it no longer becomes a cache, it becomes your primary storage. Like anything else that's new and disruptive, there are certainly challenges that we as suppliers and the industry is facing with things like usage models being a little bit different in technical requirements than your historical data storage would present. Interfaces are certainly varying. We've got things like AHCI, NVMe, SOP as interfaces that are being talked about. Certainly, form factors is another one that are kind of being defined on the fly. Right now, for example, we have at least four different form factors that are being designed at Micron just to service the PCIe TAM. Next slide, please. If I take a look at Micron's approach to servicing this market by interface, let's start with SATA.
You just heard Justin talk about what we're doing to service the client SSD market. For SATA in the enterprise, we are building upon that foundation. Reuse for us is something we take pride in. We do this by adding enterprise features such as things like data path protection, different over-provisioning, et cetera, to service this segment. Right now, we have 2 generations servicing this market, the P300, which was introduced back in 2010, the P400e, which was introduced 2011, and we're currently in qualifications with OEMs on our latest P400m drive. Each of these drives continues to build on our client investment, again, with the addition of features and new lower-cost NAND lithos required to meet our customers' demand. The P400m, which, as I said Qualification with OEMs today utilizes our 25-nanometer enterprise-level MLC NAND. Next slide, please.
As I said, I want to spend a little bit more time talking about PCIe, because again, I think it's one of these sub-segments that's certainly new, exciting, and I would argue the competitive landscape is certainly going to be heating up in this particular sub-segment. I want to start by talking about two very different architectural approaches to this segment. At the highest level, I look at it as being two fundamental approaches to building PCIe cards. The two that I want to talk about are what I would call host-based management versus drive-based management. Simply stated, what that means is whether or not the server itself manages the card or does the card manage itself. These are very different approaches that, frankly speaking, yield different results. Let's start with what our competition is doing. It's what I call host-based management.
What this does is it pushes the management of the drive back onto the host or, in most cases, the server. It uses the system CPU and system memory to manage the drive. Now, what's nice about this approach is that you can use a smaller FPGA on your drive, and it allows you to do relatively quick time to market. You pay a penalty for this approach. The penalty that you pay for this approach is that you give up performance, most importantly, the drive ends up utilizing valuable system resources to manage your card instead of doing what you bought the server for, which is compute work. This is the approach that our competition has taken. Our approach, again, is quite different. It's what I call a drive-based management approach.
This approach allows the drive to handle the management itself, freeing up system-level resources, again, to do what the server was purchased to do, which is compute. The downside of this approach is that it requires you to develop your own ASIC, and manage it yourself. For all that hard work and investment, you get a much higher performance drive, as I said, you don't use valuable system resources. Just to illustrate the difference in performance that you get, our P320h drive that we're shipping today has more than 2x the performance of our competition. We're delivering random read performance north of 800,000 IOPS and random write performance somewhere close to 200,000 IOPS. Again, this is roughly 2x our competitor, which in this market is huge when you're looking at the total cost of ownership in terms of $ per IOP.
I fundamentally believe that as customers evolve their understanding of the value and usage of high-performance PCIe drives, that this will be the winning architecture. I think we're heading in the right direction because of this. The other two things that I mentioned earlier are form factors and interfaces, I want to talk a little bit about form factors. As you look at this slide, we developed a unique form factor, unique to PCIe, not unique to the industry, with a two and a half version of this drive with our partner, Dell. It allows it to fit in an existing socket and improve serviceability because it's in the front and not inside the server. As you can see, we also worked with one of our other partners, EMC, to provide them with a caching solution for the front of their storage devices.
Again, somebody that values high-performance PCIe values the approach that we've taken to this market. Next slide, please. Let me finish the interface discussion by focusing on SAS. Given that SAS is growing at about a 36% compound annual growth rate, it should be no surprise to anybody that Micron has its eye on this interface. We're building upon our SATA and our PCIe system knowledge, and most importantly, our OEM relationships, to develop world-class SAS products. We are currently in qualification today with OEMs on our first SAS drive called the 410M. The six gig SAS drive uses our 25-nanometer MLC NAND technology. We're really excited about this and have received very positive feedback from our OEMs with regards to the performance of this drive.
As we look to the future and see SAS users admiring the performance that we get out of our PCIe in a 25-watt envelope, my expectation is that we will potentially start to see pieces of PCIe showing up in SAS and vice versa. Out in time, it wouldn't surprise me that in the 2015, 2016 timeframe, that some of these architectures start to consolidate around a higher power, higher performance interface that allows you to take advantages of both the legacy world and, frankly speaking, where the world is going. Right about this time will be interesting because this is when we expect to be really ramping some of the new memory technologies that we think will put us in a leadership position to take advantage of our investment in solid-state devices that we're making today. Next slide, please.
Finally, I wanted to end with, again, a very brief summary of what I think enables Micron to win in this space. I think as you heard from both Glen and Justin, this concept of being a vertically integrated manufacturer is a very important piece. We have the luxury of calling folks on the phone here at Micron and having very detailed technical conversations about
the capability of the current silicon that we're utilizing in our drives, but most importantly, what we can do in future designs at the silicon level to make the most out of the drives that we build. Having that vertically integrated manufacturing infrastructure that we're building upon, I think, is critical to us being a winner long-term in this space. I think, as you've seen, we have a very targeted portfolio of enterprise products. SATA, we're building on our investment in client. PCIe is a homegrown solution that puts us in a performance leadership position, and we expect to build upon that with our current OEM relationships.
Likewise, in SAS, we're building upon the relationships that we've developed through the client-level business and into the enterprise business to move and to ship our high-performance, high-value products into OEMs that value that vertically integrated solution from a manufacturer such as Micron. To sum it up, I think we have a great product portfolio. We have key relationships with both storage and server OEMs, and we're excited to be in this business. With that, I'll turn it back to Ivan and Glen.
We will now take questions from callers. Just a reminder, if you're using a speakerphone, please pick up the handset when asking a question so that we can hear you clearly. Operator, go ahead and take some questions.
Certainly. Again, ladies and gentlemen, if you do have a question you may press star followed by the number one key. Our first question comes from the line of Shawn Webster from Macquarie.
Hi. Hey, great. Thanks, guys. Thanks for hosting this call. It's very educational. A couple of quick questions. One is, I was curious on what the mix today is for your business on SLC versus MLC and how you expect that to evolve over the next couple of years.
This is Glen, Shawn. I'll go ahead and take that. Overall, the vast majority of what we ship is Multi-Level Cell. We do have a pretty good business still with Single-Level Cell into some very high reliability, mission-critical applications. We do run Triple-Level Cell as well, in fact, we've recently increased our production of Triple-Level Cell. Essentially, our strategy here is that with the customers that we're engaged with, they tend to value some of the performance and reliability aspects of our Multi-Level Cell devices. We've received very positive feedback from our customers that are using these devices in some critical applications. Our strategy with Triple-Level Cell is that we run it to the extent that we really need to adapt to market conditions.
Given where the NAND market is today, we've decided to increase that, and we're currently running in the low single digits. We can modulate that, again, depending on the amount of capacity we need to serve in sectional portions of the segment that aren't really a strategic focus for us, but it does become important during softer industry times like we're currently faced with.
Okay, if I do some simple caveman math for the units and the revenue numbers you gave us, I get an implied I know you don't guide for the future pricing, but it looks like you're looking at 20%-25% a year price reductions. Given the challenges in terms of migrating to future process nodes, is there any barriers, or do you think they're seeing those kind of price reductions as we move to the mid-teens or even lower in terms of process node technology?
Great question. It's certainly one that receives a lot of attention within Micron these days because it is getting so much more difficult for us to get to that next lithography. However, we're pretty bullish on our technology roadmap. In some of our other discussions that we've had publicly, we've talked about the opportunity that we have to scale our existing planar NAND technology. We've also talked about opportunities that we have to go vertical with the so-called 3D NAND technology. We also are placing a couple big bets on some emerging technologies. We're confident that between that multi-pronged attack, so to speak, that we're going to be able to keep up with the delivery of these market cost reductions that enable new inventions for the future.
Thanks. Just one final one, and I'll go away. On the controller technology side, that's clearly very important, maybe even getting more importance going forward. In terms of your mix today within solid-state drives, how much is Micron-designed, Micron-only controllers? I was wondering if you can give me a sense of what the share of the bill of materials in a solid-state drive is for the controller part. Thank you.
Yeah, you bet. Maybe I'll answer it, I'll turn it over to Justin here for further clarification. I think that our best proof point for our internal controller capability is on that P320 product that Ed Doller just referenced. That sets a pretty high bar in the industry. That's probably the most difficult SSD or SSD-like product to produce, and that's using a controller that has our firmware. Of course, we use partners to manufacture that ASIC because we don't do that within Micron's fabs. That is an internal design that we're pretty proud of, and I think it's a very good proof point for that internal controller capability. That said, our product portfolio variety is so wide, we know that we're going to have to partner with others in order to be successful in this market.
In our client SSDs, that's certainly where we've been doing quite a bit of that. I'll turn it over to Justin to answer the rest of your question there.
Yeah. I think, as Glen was saying, for client SSDs, we're partnered on all of our controllers today. None of those are developed in-house. What I want to expand on that is while the controller is important, the firmware is actually even more important in our mind. While we are partnered, it doesn't mean that we're just taking an off-the-shelf product. I think it's pretty widely known that we're partnered with Marvell today, and that's been a good partnership, so we're using Marvell's hardware. We work closely with them to get the features we need in the product to make sure that it works well with our NAND, we develop all of our firmware 100% in-house. That really sets us apart and allows us to not only design in very high quality, but also be very reactive to our customers' needs.
The second question, I think you had a question on how much of the BOM was controller-based. Is that?
Yeah, just a ballpark.
Yeah. I guess I'll expand that and maybe answer it in percentages. The controller is a piece of what goes into the SSD besides the NAND, but I think what you're getting at is how much of it is NAND and how much of it is not, and it really depends on the capacity point. For primary storage in the 128 gigabytes and above range, that will range from 10%-20% of the total cost.
Thanks a lot.
Thank you. Our next question comes from the line of Monika Garg from Pacific Crest Securities.
Yeah. Hi, thanks for taking the question. Just as a follow-up, first of all, to the last question, could you talk about your controller strategy on the Client side and Enterprise side? Client side, I think you just clarified, seems more kind of collaboration with external parties. How about on the Enterprise side? Is it all internal or again on the collaboration side?
Yeah, I'll answer that. Thanks for the question. I would say right now we have a kind of a dual-prong approach. First and foremost, on the low end of the Enterprise market, as I stated during the discussion, we will utilize the controllers that Justin uses in his Client drive, and we will build upon those, making sure that we have the right features for a particular segment in the Enterprise market. For the PCIe, the controller was developed internally. For our future products, if you look at our PCIe and our SAS roadmap, we will have another version of an in-house developed controller for SAS and another version of an in-house controller developed for PCIe. The controller that's being developed for SAS will be able to be reused for a SATA.
Again, if I look at the Enterprise segment from a SATA standpoint, there's a couple of different usage models. One of them, again, is a boot drive application, so to speak. I think the strategy of reusing Client with the appropriate features works very well there. Clearly, there are customers who require features, things like security for data at rest, et cetera, that we believe will fundamentally be better serviced off of the controller that we are designing in-house for our SAS segment.
Okay, thanks. Is it possible to share the BOM % on the Enterprise side also?
I actually prefer not to. Well, let me just comment on this. I would say that the % is certainly lower, and it's primarily lower because of the densities that we're shipping into Enterprise are typically higher than what we're shipping into the Client side of things. I would say on average, it's lower than what Justin stated.
Sure, definitely. Another question on, you talked in the beginning of the presentation about integrated software hardware appliances and the Virtensys acquisition. The question is, when are you planning to introduce the product in the market? The strategy would be to kind of go with the storage OEM like EMC, which you are doing the VFCache solution, or are you planning to go completely Micron solution?
Great question. A couple things with regards to the Virtensys acquisition. Number one, Virtensys had a virtualized appliance in the market that we continue to support, and we will be seeing revenue from that by, I would say, by the middle of next year to a very select handful of customers. That is a virtualized I/O appliance without storage. Obviously, we bought the company because we value their PCIe sharing technology. What you will see from us and what we are actually demonstrating right now with a handful of select OEMs is, in fact, that appliance with our P320h PCIe card sitting in that appliance. The value of having a PCIe card in an appliance is twofold. Number one, we can get much more storage into an appliance than we can sitting in, for example, a blade server, right?
Blade servers have very limited space. Today, you have to open up a blade and you put less than a terabyte of storage in them. Once they're in there, they're native to that particular blade. What we like about this solution is, A, as I said, we can get much more storage into the appliance. Number two, once we hook these up via PCIe interconnect to either a blade or a rack server, you can partition the storage and you can do it on the fly by sharing the storage where it needs to be shared, which we think, number one, has value. Number two, what it allows you to do is not only split up the storage dynamically, but allows you to fundamentally share that storage.
Think in terms of a rack of servers, where today you may have an instance of a database sitting on every one of those servers utilizing precious storage. What happens if you put that database in a rack appliance and you have the ability to share that across each one of those servers connected to that appliance? Then your total cost of ownership starts to look dramatically better. We fundamentally think there's a large value there. As I said, we're demonstrating that with key OEMs right now. My expectation is that we will start to see
Production-level versions of this later next year.
All right. Thanks, Ed. Just one more question on the client side. When you talked about 60% attach rates in client side in 2015 in the presentation, was it just pure SSDs or it was pure SSDs plus cache SSDs included?
That was pure SSDs plus cache included.
Is it possible to share the split or?
Yeah.
Like how-
I don't have the split off the top of my head.
Okay.
Out in time, we definitely see the majority of that shifting towards primary storage.
All right. Thank you so much.
Thank you. Once again, if you do have a question, please press star followed by the number 1 key. Our next question comes from the line of Glen Yeung from Citi.
Thanks, guys, for hosting the call. Kind of a higher-level question. When you guys are out in the market competing with your SSD solution, what do your customers say are the key differentiators for you or for your competitors for that matter? What makes them choose a Micron over someone else?
This is Glen. I would be willing to bet that Justin and Ed can also elaborate on some things here with their specific segments. The thing that I hear most often is that customers want to work with us because we are fully integrated, because we have the core capabilities from silicon leadership all the way to systems leadership, and we're even pushing farther vertically, as Ed just described with our appliance strategy. Increasingly, I'm hearing from customers that that's becoming a bigger part of their decision process, and not just because of what I think has been largely talked about in the industry lately about the abilities to engineer solutions or between the silicon and the system. As we ramp into high-volume production, what's becoming very much on everybody's mind is the consistency of supply and the quality of that supply.
Certainly, as the NAND market goes through its cycles, the NAND manufacturers having that captive supply can bring something to the table that a lot of the other companies can't. That's essentially what I'm hearing. I think also in some of the successes that we've had within both the client and the enterprise segments, our ability to work closely with customers, to form strategic relationships, and do something different for them that our competitors can't, is something that's getting more and more of our attention and focus. Justin or Ed, would you guys like to add something there?
I think you covered it quite well.
Glen, I think that makes perfect sense. I agree with what you said. I think, if you look at our competitors that don't have their own NAND supply, there's clearly worry about supply, especially if that particular competitor is competing with a NAND manufacturer such as us. That's definitely something that we hear concerns about from our customers. I'd also add that our competitors that have to purchase NAND, they're clearly doing the purchasing from multiple supply points. At the end of the day, as much as we look at NAND as being a commodity, not all NAND is created equally.
What that fundamentally means is that if you're going to design a drive to utilize more than one supply of NAND, that in itself presents challenges above and beyond what I think we have to do with the intimate knowledge that we have of our own NAND.
Maybe to bring that a little bit nearer term, in the press now, they're talking about supply reductions by some of Micron's major NAND competitors. Is that the type of inconsistency that you're talking about, or is it really inconsistency in terms of a bigger supply condition?
Yeah, no, I think that's a real good point. We're with our industry peers right now in that we're not real excited to add a lot of capacity either, and that's going to remain so for some time. As this long-term demand and these new applications kick in, you bet things are going to tighten up once again, and I think that's going to lead to a different dynamic that will favor the NAND suppliers who are manufacturing their own SSDs.
Okay. One last question, which is just about the controllers again. First of all, can you just, for educational purposes, differentiate between an SSD controller and a NAND controller, a NAND-specific controller, and then talk about where you think the value is between the two of them to your customers. Where is greater value?
Well, let me just maybe take a step back here. When people, I think, use the term NAND controller versus SSD controller, I think it's two points along a continuum of complexity. All controllers must somehow manage the NAND media. Again, as the technology scales, that alone gets very difficult, and I think that's something that certainly we have an advantage in. The other side of that controller is what talks to the outside world, and there's a big difference between an SD controller or a USB controller versus a client SSD SATA SAS controller versus a PCIe controller. Each of those are very, very different. The key, and what we're doing here at Micron is we're developing our controllers in a modular way with these building blocks where we can rapidly and efficiently mix and match them across those different product families.
Is that something that you would characterize as a differentiating factor for Micron? How do your competitors try to do the same thing, or can they?
Yeah. I think our competitors, depending on who you classify as a competitor, will certainly have a different strategy. If you're a USB or an SD company, you're going to have a very narrow focus strategy. I think for us, where we have a full product portfolio, it's in our best interest to really create very reusable, efficient platforms. A lot of the learning that we have in different segments we can easily apply to the others. As Ed mentioned earlier, we're very interested in reuse. I think for a company like Micron, that's a very natural approach, and I do think it is one that we have a competitive advantage in because we've grown our capability through the ranks, so to speak.
As Micron first entered the NAND business, a lot of the focus was on the more traditional markets, USB, SD devices, we had to hone our controller skills, based in those markets. As we've climbed the value chain, we've carefully expanded that capability. As Justin mentioned, as we first deployed the client SSDs, we took advantage of the fact that we have our own channel and initially started our ramp into high-volume manufacturing through our Crucial capability. Once we gained confidence in our controller and our manufacturing capability, we were able to engage with OEM and go through the hurdles of some of the intense OEM quals that you have to do to engage such markets. As Ed just demonstrated, we've hit an even higher hurdle of going through those same qualifications with some large enterprise OEMs as well.
It's been kind of a process for us, and we do think that we've arrived at a pretty good point that does give us a competitive advantage with a very solid controller strategy.
Good to hear. Thanks a lot, guys.
Thank you. Our next question comes from the line of Kevin Cassidy from Stifel Nicolaus.
Thanks for taking my question and thanks for giving this presentation. With your SSDs, what's the split right now for Single-Level Cell devices versus Multi-Level Cell, and what do you see for Triple-Level Cell in the future?
Okay. Yeah. I'll let Ed and Justin both answer that question, because I think it's a different answer for those respective markets.
Yeah. I'll go ahead and start with the client. For the client space, today it's 100% MLC in our client drive, and really no demand or desire for an SLC product in the client. MLC provides all the performance, endurance, reliability that we need for the notebook market space that we're addressing. Looking forward, I think there was a question there on Triple-Level Cell in SSDs. That's not something that we have in the market today. There is interest there. It's an area we continue to investigate. Really, without divulging any future product plans, I think I'll leave it at that.
Yeah. I think for enterprise, definitely a bit different. I would say right now we're probably in the 70% MLC, 30% or so SLC. The reason for SLC is primarily because of the usage conditions associated with something like a PCIe caching solution. As you can imagine, if something's a cache, the number of writes per day, fills per day, total bytes written, pick your reliability metric, is quite different in that environment. For us to meet some of the very high bars that are being set by our customers' usage models, we have SLC offerings on those interfaces.
Okay. Just to make sure I'm clear on this, is that a different controller than for an SLC than would be for an MLC?
No. Our controllers are designed to be able to deal with both of them. Primarily, the fundamental difference between handling both of them has to do with the ECC engine that we embed into our controllers. The reality is you could always hook one up. It's just a matter of what the reliability will look like as you run it through the error-correcting engine that's put onto that controller.
Okay. Maybe, we didn't talk a lot about the process technology as you go down to 1x level. How much does the controller change there? Also, I know in the roadmap, there's plans for going to 3D, I just want to know what happens, again, with the controller technology as you go to these different process nodes.
Yeah. Let me comment on that first, then maybe Glen or Justin can add something. It really fundamentally comes down to understanding what the usage conditions are of the system that you need to deliver. There's many ways to deliver those. You can do that hopefully by starting with the best ECC engine you know how to do, I think that's the general approach that we take, there's a lot of knobs that you can put into a system. Again, I think given our understanding of what the NAND does down to the electron, I think we're in a fantastic position to know which knobs we can turn and how to optimize those knobs to deliver the highest performance, highest reliability at a particular cost point in the market.
What that means is when we get to the 1x nodes and beyond, we know the knobs that we can use, I fundamentally believe that we will be in the best position to deliver that combination of reliability, performance, and cost to our customers as we continue down the scaling of our NAND technology.
Okay, great. Thanks.
Operator, I think we have time for one more question.
All right. Our final question comes from the line of Ryan Goodman from CLSA.
Hi, thanks for taking my question. Question on the caching solutions. You had introduced a product called Adrenalin back in January. Now it sounds like you're not looking for a lot of growth in this sub-sector in the market, and it sounds even like you're de-emphasizing this in a longer-term strategy. Just wanted to know if you could comment on that and maybe talk about how you're going to approach this market in the next couple of years.
Yeah. Excellent question. I guess I'll start with Adrenalin is a cache. It's offered in the market as a cache, but I don't want that product to be confused with the caches that are being used in Ultrabook. The Adrenalin is a 50-gigabyte cache. It's a higher capacity. It really was designed and brought to market as a high-end cache, a main storage cache. Because if you look at a 50-gig cache, you can really put just about everything you're using on a daily basis in that cache.
Really what you have is an SSD for all practical purposes, along with some caching software that allows you to attach it to your hard drive that, this is offered to the Crucial market, probably the person buying that has a hard drive already, and they're upgrading. It allows them to attach into that hard drive, continue to have terabytes of space, and all the benefits of an SSD. That's really a different approach from what's happening in the Ultrabook space, where you see 24-gigabyte caches, and in that space, it's really all about for as little money as possible, how do I get some of the goodness of an SSD and hit Intel's minimum target requirements?
Through the upgrade market and the e-tail space, we'll continue to focus on that Adrenalin product and continue to have offerings like that, which are really, we're taking and we're bundling caching software with would really be considered a main storage SSD and allowing somebody to take a lower capacity main storage and have the benefits of the hard drive space behind it.
Okay. That makes sense. One other question, going down a different line. In Ultrabooks, it sounds like Intel is pushing more for a standardized form factor. They have this next-generation form factor. In the market today, we're seeing a variety of form factors coming out, especially from some of your competitors. Just wanted to know your thoughts on that and the strategy there longer term as well. Thanks.
Yeah. I can't really comment too much on Intel's desires for the Ultrabook market other than what they've made public already. I think you're going to see different customers try and—or different OEMs try and differentiate themselves from each other. You will see different offerings in the market, whether that be a transformable, where it's got a touch screen and a flip screen, or maybe a screen that removes from the keyboard. I think you're going to see a lot of innovation in this space as the OEMs try and figure out really what best addresses their customers' needs.
Sorry, I just meant in terms of the actual SSD form factor, though.
Oh, the SSD form factor? Predominantly what we're seeing today is mSATA in these devices. There's definitely discussion about some other module form factors going forward, but really, they all kind of address the same desire, and that's to get an SSD into a smaller form factor that enables a thinner and lighter Ultrabook. To us, we're pretty ambivalent about what that module form factor is other than we like standards because it's just much better for the industry to have standards.
Okay, thanks.
With that, I would like to thank everyone for participating on the call today. If you'll bear with me, I'm going to read the safe harbor protection language. During the course of this call, we may have made forward-looking statements regarding the company and the industry. These forward-looking statements and all other statements that may have been made in this call that are not historical facts are subject to a number of risks and uncertainties. Actual results may differ materially. For information on the important factors that may cause actual results to differ materially, please refer to our filings with the SEC, including the company's most recent 10-Q and 10-K. Thank you.
Ladies and gentlemen, thank you for participating in today's conference. This does conclude Micron's