Good day. My name is Sean. I will be your conference facilitator today. At this time, I would like to welcome everyone to Micron Technology's SSD Market 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 key. If you would like to withdraw your question, you may 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.
Thank you. Welcome to the Micron Technology SSD Market Update. On the call today is Glen Hawk, VP of Micron's NAND Solutions Group, Ed Doller, VP and Chief Memory Systems Architect, and Allen Holmes, VP of NAND Solutions Group Marketing. This conference call, including audio and slides, is also available on micron.com. Our call will be approximately 60 minutes in length. There will be an audio replay accessed by dialing 404-537-3406 with a confirmation code of 94749756. This replay will run through Thursday, August 23rd, 2013, at 5:30 P.M. Mountain Time. A webcast replay will be available on the company's website until August of 2014. 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.
Hello, everybody. Again, my name is Glen Hawk, I'd like to start off with a brief introduction before turning it over to Ed and Allen. To start off here, I'd like to put things in a little bit of perspective. There are a lot of reasons to be excited about the memory industry right now. Some of you may be excited about the supply-demand tightness that's in the market right now. Others may be excited about the consolidation that's occurring, much of it in favor of those who actually make memories or make NAND in particular. Others may be excited about the plethora of new startups that are out there that are applying flash components and flash SSDs in ways that previously were not really envisioned. Those reasons excite us, too.
At our core, we're a technology company, one thing or one overall trend that really excites us is that the memory technology innovations that we're driving here at Micron are enabling innovations around the world that change the way people live, change the way they work. These changes in the way people behave and act are creating new devices, new demand for devices that in return drives a larger thirst for more memory. That encourages us to push harder and invest more deeply for some of those innovations. It's a very nice upward spiral that we find ourselves in. A lot of the data and information that we're going to talk about throughout the rest of this afternoon is going to be a little bit more about the here and now.
Please keep in mind that we firmly believe here at Micron that memory is changing the world, we really think the best is yet to come. If you think back over the last few years, maybe 30 years ago, people and analysts probably wouldn't have projected the number of gigabytes required to support the transformation of digital audio and video. If you go back maybe 20 years, probably we weren't very good about predicting the amount of bits that would be consumed by mobile platforms, smartphones in particular. If you go back just 10 years, I'm not sure that we were really that great at predicting how much memory would be, and how much solid state memory in particular would be consumed by cloud computing.
Those are all trends that have happened, we're well aware of, we'll talk about the rest of the day, but I'll assert that there's another wave yet to come that few of us can imagine, and we won't talk a lot about that today. If we do talk about the current outlook on the next slide, what we see is we firmly believe that there is continued strong growth for NAND. What we're showing here is the 2013 to 2017 projection. As you can see from our source here, we're using Gartner. Many of the analysts have similar projections. We're all seeing very strong, consistent growth over the next few years. Starting from the bottom up, we have some of the traditional NAND segments that really fueled some of those early innovations that I mentioned earlier.
Removable storage, industrial, medical, automotive applications are still an important part of the market and our business. We see tablets here, a relatively new entrant. We see mobile. What we're going to talk about for the rest of the presentation here is the real breakaway segment, which is SSDs. We've been focusing quite heavily on this recently, and we'd like to give you an update on that. To pick that apart a little bit and look at what's happening within that solid-state drive segment, at a high level, there's two components of it that are important. They're similar, but they're similar and they're important for different reasons. What we're showing here is the difference between enterprise solid-state drives and client solid-state drives. We're showing both the gigabytes that are projected to go into those segments, as well as the dollars.
By comparing the two, you get some feel for the differences. On the left, by looking at the gigabytes, you can see that the capacity or the gigabyte growth is really driven by client SSDs. That makes a lot of sense. There are certainly a lot more client SSD devices in the world than there are servers and storage array networks, et cetera. You can see by a wide margin, the vast majority of the capacity is required to fuel that client growth. That's very interesting to us because at Micron, scale is important, and we know that it's important to participate in that client segment, we certainly have a focus there. If you look at the right and you look at the partition of dollars or revenue between enterprise and client, you see a very different picture.
You can see that it's relatively equally balanced between enterprise and client. Of course, what that means is that there's more value there in the enterprise segment. It's much more difficult to serve. There are a lot of core competencies and capabilities that a enterprise requires. What attracts us there, of course, is that higher value, those higher margins, and the more stable margins. They're rather counter-cyclical to client SSDs and a lot of the other segments that we serve with our NAND businesses here at Micron. As a result, that's why we have been on a multi-year campaign here to move up the value chain from silicon to systems. We built upon our silicon technology leadership and our NAND product leadership, we, over the last few years, have been investing very heavily in some of the key elements to be a leader in solid-state drives.
That includes internal and external controllers. It includes a focus on both client and enterprise because we see them as being very complementary to one another and important parts of our business. We do have our eye on the next level of integration above that, which is integrated software and hardware appliances. Before I turn it over to Ed here, I would like to make one note that, even though our focus is on climbing that value chain, by no means have we lost our tenacious focus on our silicon technology leadership. Earlier in the week, we did announce that we are sampling our 16-nanometer NAND technology. An image of our 128 gigabit MLC device is shown here on this slide.
Some interesting aspects of that are we can get nearly six terabytes of storage on a single wafer, and we do plan to introduce some of our solid-state drives using this very same piece of silicon, in early 2014. Once again, our focus for today is on solid-state drives. At this point, I'll turn it over to our Vice President and Chief Memory Systems Architect, Ed Doller.
Okay. Thanks, Glen. I think as Glen stated, both NAND technology and capacity are the foundation of SSDs, and having both is enabling us to develop a complete market-leading storage portfolio. We spent the last couple of years focusing on building a complete product portfolio. By complete, I'm referring to interface complete, form factor complete, and of course, market segment complete. What you're seeing here is really the result of our focus. I'd like to spend maybe a minute or two pointing out a few products and their relevance to the market. Starting at the top, we have a P320, which is the highest performing PCIe drive in the industry, with bandwidth that saturates the interface, has sub 50 microseconds latency, and well above two gigabyte per second sequential write bandwidth.
When you achieve that level of performance in an SSD, it frankly opens up a world of new uses, and we're starting to capitalize on those new uses with offering drives to take advantage of them. We still have new customers that frankly can't believe the level of performance that these drives actually deliver. Having PCIe SSDs in a two-and-a-half-inch form factor is actually changing the way PCIe storage is thought and, of course, utilized. Next is our P420. That drive builds on the success that we've had on the P320 and adds capacity and cost improvement over the 320 by trading off performance and fills per day. For those applications that don't need the high performance and high fills per day, the 420 is a great solution for that market. Next, you move down to our enterprise-class SAS and SATA drives.
We launched these drives several quarters ago and continue to receive incredible market acceptance of these products. Next are our products designed for segments that require less fills per day than classic enterprise, but still demand high reliability and high performance. The P400e and the 550 are doing a great job covering this segment. Last not least is the M500, which utilizes our 20-nanometer industry-leading NAND and was the first one terabyte SSD in the industry. I can tell you, keeping up with the market demand for this drive has been no small feat. If you compare the product portfolio to others in the industry, you can see that Micron product breadth is second to none. Let me take a moment to point out the column to me that is by far the most important on here, and that's the NAND column.
I think because NAND makes up the majority of the bill of materials, we fundamentally believe that it's the necessity to success in this market. I'll go as far as to say that there is and will continue to be an inexorable move of market segment share to those with it and to those that have product portfolio breadth to utilize it. We have both. Before I turn it over to Allen to talk more about the details of our current business, I think it's important to recognize that customers, both storage and compute, see Micron as a company that they want to do business with. It's the combination of having a solid foundation that starts with NAND technology capacity and ends with the broadest SSD portfolio in the industry. With that, I'd like to turn it over to Allen for a business update.
Great. Hey, thanks, Ed. I'd like to spend a few minutes kind of talking through how our business is really shaping up for SSDs. In particular, I'll start off with the client piece. If you can move to the next slide, please. Ivan, it's interesting to note that it took us about four years to really ship the first 1 million units for Micron. Today, we're actually shipping about 1 million a quarter. I think that's, again, a testament to what Ed was saying in terms of having our broad portfolio and really building out our business. If you look at how our customers are really starting to adopt the technology, we think there was an interesting trend that happened last year.
If you look at around the Q1 2012 timeframe, the pricing and the market that Glen was describing was very advantageous for customers to start to really adopt SSD technology. In particular, we really saw growth in our business, as you can see from the graph. In particular, we're growing 2x faster than the market. I think that's something that we believe is, as Ed mentioned, having access to the technology, having access to NAND capacity is certainly in our favor. One thing I will note is, in many of the SSD suppliers, the average density that's being shipped into the industry is really around about 128, 129 gigabytes. For us, we're shipping somewhere around just south of 200 gigabytes.
That's an interesting trend for us, because as we're driving, again, higher density SSDs into the marketplace, we're finding ourselves into some very unique market segments and opportunities. We'll continue to grow on that over time. If we move to the next slide, let's talk a little bit about enterprise. We're really excited about the trends here. Specifically looking at, as Ed pointed out, as we've grown out our portfolio, the number of market opportunities and the number of customers we engage is growing every day. As we look out over the horizon from year-on-year growth from Q2 2012 to Q2 2013, we're really kind of trending around 120% growth over that horizon.
Again, as we build out our portfolio and we engage with enterprise customers, knowing that the qual time is a bit longer than our client business, it gives us the opportunity to see future growth. One thing I'll note here is the density that we've seen in enterprise has been low, what you'd expect. Average has been around 200 gigabytes. Today, we're just slightly ahead, around just 220 gigabytes. We continue to see demand and opportunities for in excess of a terabyte, we'll continue to look for opportunities to service that. If we move to the next slide, really just want to talk about our SSD growth. Right now, it represents about 25% of our revenue, overall NAND revenue. We continue to see this a benefit for Micron, of course, adding more value, and as Glen mentioned, moving up that value chain.
We believe that's a competitive advantage for Micron. As we look at how we're leveraging our portfolio, how we're engaging with new opportunities, with new customers, we really view this as something that we'll continue to take a look at. The next slide, I really want to talk about how we're engaging with storage in general. I think it's interesting to look at how we're servicing that market. About 50% of our NAND is going into the storage markets. I think this is a focus area for us, and we continue to look for ways to increase our footprint with Micron Technology into the SSD marketplace. Okay. If we look at, next slide, please. From a summary perspective, as Glen pointed out Opportunity. We're certainly laser-focused on taking advantage of that opportunity.
As Ed stated, we have a very broad portfolio, we intend to continue to develop and invest in maintaining this portfolio. We are growing faster than the market, we intend to continue to leverage our NAND and the technology that we're developing. We're looking at, again, half of our business is going into the storage industry, we continue to expect that to grow. We know, as Ed pointed out, we're just seeing a number of customer OEM opportunities to align with Micron and get access to our technology and our NAND capacity. With that, I'll turn it over to Ivan for Q&A.
We'll now take questions from callers. Just a reminder, if you are using a speakerphone, please pick up the handset when asking a question so that we can hear you clearly.
Ladies and gentlemen, on the audio lines, simply press star one on your touch-tone keypad. You may remove yourself from the queue by pressing the pound key. Again, if you have an audio question, simply press star one at this time. Our first question comes from Stephen Chin of UBS. Please go ahead with your question.
Hi. Thanks for taking my question. Can you hear me okay?
Yep, we can hear you.
Great. Ivan, I was wondering, in terms of the percentage of your NAND supply that you're willing to allocate towards solid-state drive products, is there a natural ceiling or a limit that you guys would target, or is that something that you will take as it comes? Secondly, in terms of the competitive landscape today, just given the tightness in the supply, are you seeing the market as being more competitive, just in that the companies with capped supply are gaining share right now, or is it less competitive just because you have fewer players, like smaller companies that can't compete because they don't have the capped NAND supply?
This is Glen. I'll take the first part of that question, then Alle n, I think, can dovetail into some of the latter parts of your questions dealing with some of the current competitive phenomena that we're seeing. I think your first question was, the one I'll address, which was, do we have a ceiling, so to speak, on the amount of our capacity that we're willing to allocate toward SSDs? Let me add a little of our perspective to that is that even though SSDs is one word, so to speak, we certainly don't see it as one segment. Solid-state drives are finding their way into this incredible diversity of segments themselves. At a high level, we talk about enterprise and clients, each of those can further be subdivided into tremendous complexity.
We think that those sub-segments, so to speak, again, can be somewhat counter-cyclical of one another. On the enterprise side, at a high level, there's server-side flash and there's storage-side flash. Within what we call a client, there's not just clients for corporate clients, things that large corporations buy to make their employees more productive, there's consumer clients, typically more price-sensitive devices, might not have the same security concerns as the corporate clients. The consumers and the corporations oftentimes find themselves in different cycles. The refresh rate among corporations is often very different than, say, the holiday season purchasing cycle for clients. We can subdivide that even further. We're seeing a lot of exciting opportunities for solid-state drives in other segments. Like in automotive, for example, one could argue that the future automobiles are nothing more than tablets that you can ride in in the future.
The amount of flash and the use of SSDs is a very exciting opportunity there. There are a number of other devices as well. My answer is, we certainly don't have an artificial ceiling just because of the end product that we're selling our NAND into. We look more carefully at the end segments, the end customers, and make sure that we're comfortable with that diversification. We're not hitting any limits by any means at this point in time. With that, I'll go ahead and turn it over to Allen.
I think from a competitive tightness or the impact it's having on the landscape, clearly, I think others in the industry have signaled that they've seen limited NAND supply. I think that's clearly one of the advantages of having access to the capacity and really modulating it for the market opportunities that we see. I think that we're certainly got a little bit of tailwind in terms of market opportunities that we're being exposed to. I think the growth in some of the examples that Glen pointed out are creating some tightness. Again, I think being able to navigate through that and modulate our capacity is certainly a competitive advantage for us as it relates to where we want to put our SSDs in terms of customers and segments.
Okay, great. Just one last one if I could. Glen, just given your comments about how you guys are approaching this more increasingly from an appliance or a complete solution standpoint, how should we think about the margin profile for these enterprise-type products? Are there any specific companies that you're benchmarking yourself against as you create these solutions and think about pricing or margins for the products? Thanks.
Steven, we certainly think about that quite a bit. It has been new frontier for a lot of what we've historically done here at Micron. Of course, when we look at solid-state drives, we see the value add from just simply creating a system-level product in the first place, and then of course, we see the value add with a lot of the other product aspects that really make that piece of hardware, so to speak, a total solution. It can be the software that accompanies it, sometimes in the form of tools, drivers, et cetera, manageability-type software. The services and the other support that we provide are very important to certain customers. The ability to walk into a customer and hold their hand and really help them design that product into their application and optimize it, we're tapping into all sorts of value there that historically we haven't.
When it comes to benchmarking ourselves, we really have to benchmark others who are doing bits and pieces of all of the above. There's relatively few of us that can participate in all of that, from the NAND silicon all the way up through the service, the software, the support, and all of those sorts of things. There are a lot of individual companies that are doing very well with certain aspects of that. We do try to take a look at all of it. I think with the chart that I'd shown previously, that clearly shows the dollars per gigabyte for enterprise being so dramatically higher than client. I think the first order, that gives you a feel for the incredible difference that we're seeing just between those two segments.
Again, enterprise does certainly rely on some of those other value-added features more so than client. Anyway, the benchmarking is certainly going on, but it's quite a process to look at an array of fellow travelers here and see how we're doing, and we're pretty comfortable with our progress so far.
Great. Thanks so much.
Our next question comes from Douglas Freedman of RBC Capital Markets. Please go ahead with your question.
Great. Thanks for taking my question, guys. We hear a lot in the market right now about some of your competitors entering the SSD space with 3-bit per cell solutions. Can you tell us how those are stacking up and whether you're going to need to offer a like product?
Ed. Let me take that question. This is something that we continue to look at. I think with our majority of our focus right now on the enterprise market, I think 3-bit per cell in the enterprise is probably something, given the requirements of the enterprise from a performance and a reliability standpoint, that is certainly something that we don't see, at least in the segments that we're focusing on in the enterprise. On the client side, I think it's important to make sure that when we compare TLC costs at the SSD level versus a non-TLC or an MLC, that you really understand that our 20-nanometer MLC is extremely cost-competitive against a TLC. What we've been doing is certainly using our 20-nanometer 500M drives to compete quite nicely against a TLC drive.
From a cost standpoint, we think we've got a cost-competitive solution, and clearly from a performance standpoint, having an MLC solution gives you the performance differential over TLC. With that said, I do fundamentally believe going forward that there's going to start to be these cold storage type of applications, where previously, most of us that have been in the memory business for quite a while never thought we'd see the day where we have cold storage that have been serviced by hard disk drives previously, looking at an extremely cost-effective solution from a total cost of ownership standpoint, that we fundamentally believe down the road is going to open up a just enormous opportunity for NAND. Glen's comment during his introduction about there's things that we never thought were going to happen. We're living through some of them right now, for sure.
I'll tell you, this concept of data centers going, now that I've got a taste of NAND, and I really understand the value proposition and the fundamental total cost of ownership firsthand, I'm looking at every hard drive I've got and looking to replace them with solid-state technology. Perhaps there's opportunity for TLC or just MLC on a more cost-effective technology node to go after that. I'm not sure we have enough capacity in the industry collectively to really put an enormous dent in it, but it's going to start hurting, there's no doubt about it.
I guess as my follow-up to that, you're talking about pushing into the cold storage space. Does that require 3D NAND, and where are you guys at with your 3D NAND roadmap?
I would first start by saying, I don't think it requires 3D NAND. I think having 3D NAND, when and if 3D NAND is a lower-cost solution than the technology we're shipping today, will certainly dictate that. I wouldn't sit here and say that the only way NAND technology is going to make it into cold storage is by 3D. I think it's a total cost of ownership and certainly a cost solution from a NAND and an SSD standpoint. In terms of our 3D technology, obviously everyone in the industry that wants to be in
The industry is investing in future emerging memory technologies, as well as 3D NAND technologies. Right now, we're not in a position to talk about what our progress to date has been. I think it's something that you'll hear more about in future opportunities.
Great. Thank you. As a reminder, ladies and gentlemen, if you have a question, simply press star one on your touch-tone keypad. Our next question comes from Monika Garg with Pacific Crest Securities. Please go ahead with your question.
Yeah. Hi, thanks for taking my question. Ed, can you maybe talk about, when I go and meet enterprise or Web 2.0 customers, I always feel that they always talk about this insatiable capacity demand in the enterprise segment. I was kind of surprised when you talked about your average capacity is about 200-220 gigabytes, because when I go and talk to them, it seems what they always complain about is that they cannot just get highest-density enterprise SSDs.
Yeah, Monika, it's good to hear your voice. I would say there's two things going on that I don't think the market has really grappled with yet. What I specifically mean by that is, you're now getting into what we call client and what we call enterprise, and this kind of gray in-between, which some of our competitors call data center, which in some case means enterprise, in some case means Web 2.0, any of these scale-out, hyperscale architectures. Your comment is right on. There's multiple usages of SSDs in data centers. I think what you're referring to is this insatiable desire to have direct-attached storage hooked up to a PCIe interface in a data center. You are absolutely right. What we're seeing is capacities ranging from 700 gigabytes up to eight terabytes.
What we've got on our product roadmap that we didn't show here today is our next version of the PCIe SSD that will have 16 terabytes raw on the component. Now you're starting to get into just this fundamental dynamic of, where do I fundamentally want to keep my data? Do I want to keep it on a server? Do I want to have a storage array network? Do I want a little bit of both? Is PCIe really a cache, or is it a storage device? I can tell you that we've seen the full gamut. We've got customers who like our P320h because they want to use it as a cache, meaning they're going to write it a lot, they're going to dump it off and refresh it, et cetera.
What you're talking about is more of what I would just call a direct-attached storage model, something that I think we all know the industry went away from many, many years ago. Yes, that appetite for NAND and SSD capacity, I think over the next two years, is going to go through the roof.
Thanks, Ed. Just another question on the PCM. Given the very, very high endurance of PCM, could you maybe talk about the role of PCM in this enterprise segment? Are you seeing some use cases or somewhere where PCM is used right now or where you think in the future it could be used?
Yeah, that's a great question. You should come to the Flash Memory Summit because I've got a keynote where I'm going to talk quite a bit about fundamentally where I see some of these emerging memories and architectures changing the way that compute occurs over time. We don't have a lot of time, but I'll give you my real quick and dirty. We've been working with several key OEMs, IBM being one of them. They published several papers, I don't know if you've got them. If not, I'll make sure that we get them to you, where they talk about the value of an emerging memory technology that has very high reliability and extremely low latency sitting on a PCIe drive. We've delivered PCIe drives to them.
They've done with us a lot of benchmarking and I'll make sure, Monika, that we get you a couple of those papers. Clearly, it's the beginning of what we see is a major transformation over the next several years of what I would call the old-school compute architecture to a new architecture that allows you to have very low latency, non-volatile memory being a more important piece of the compute platform. We absolutely see the value there. We're investing quite a bit, not only in phase-change memory and the scaling of that technology, but other emerging memory technologies that we think will help put us into a leadership position as we work with the key ecosystem players in the industry to really transform the way data is dealt with on the compute platform.
Thanks, Ed, I would be at Flash Summit, so I'll meet you over there.
Sounds good.
Thank you.
Our next-
Sorry. Next question, please.
Our next question comes from Kevin Cassidy with Stifel. Please go ahead with your question.
Yeah, thanks for taking my question. You listed down all the interfaces that you have, I didn't see NVMe. Can you just say what your plans are for that interface?
Absolutely NVMe is an interface that we absolutely plan on supporting. In fact, that 5200 drive that I probably pre-announced, and I'm sure I'm going to get some interesting email on that, is an NVMe drive. It's a controller like the controller on the P320h and the P420m that was internally developed at Micron. It's a heavy hardware-based controller, and once again, we'll be able to saturate the PCIe 3.0 interface and differentiate with, again, an internally developed controller. Yes, NVMe is critical to us, and you'll see product-level details in the near future.
Ed, maybe can you give us a handicapping of what you think of that interface going forward? Is that going to be the standard?
I think the industry right now is grappling with that. I think over the next couple of years, what's going to happen is everybody will come out with an NVMe interface. My guess is that not all NVMe interfaces are going to be the same. It's like anything else, as that interface starts to mature, yes, I do fundamentally believe that interface will be something that will be here and big over the next several years. Obviously, there's other PCIe interfaces and protocols that are being talked about. I'd say it's a little early for us to comment on that. I can tell you, as I said, that we are absolutely committed to NVMe.
Okay, great. Thank you.
You're welcome.
Our next question comes from Glen Yeung of Citi. Please go ahead with your question.
Thanks. It's a bit more of a market-oriented question, but when we think about DRAM, we often will talk about it as a % of the bill of materials, and there being some natural cap on the amount of DRAM that can end up in a client PC. I wonder if there's some similar way to think about SSD, in terms of its dollar cost as a % of client PC bill of materials.
I'll give you my take on that. Clearly, in a client, there are certainly, like anything else, you're absolutely right, there are price points by which we start to see those capacity points move up. As Allen said earlier, when we introduced that 1 terabyte drive, we introduced it at a price point, and I think we did the same thing with the 512-gigabyte client drive, where we stimulated a reasonable amount of demand. You're right, for the majority of the market, the sweet spot right now is somewhere in that 200-gigabyte range. As a user of SSDs, I'll pay every penny for it because the power, the performance, the ruggedness that it provides me personally, I think, is something that we're starting to see catch on.
I think the other thing that's driving, certainly client, is these form factors are getting to the point where, first of all, you can't even upgrade them yourself, and you can't even fit the hard drive into them. There's this concept of form factor. There's price point that are driving the client. On the enterprise, it's a completely different animal. On the enterprise, it's really a performance total cost of ownership that's driving the capacity. As Monika asked earlier, eight terabytes and beyond for these high-end drives are something that I think is critical. Maybe, Allen, you can address this as well.
Yeah. I think the only thing I would add, Ed, is that when we got to a price point for an SSD below $100, that's when we started seeing an uptick in demand and penetration into at least some of the client form factors. I think as you look at that, when you get below $1 per gig is one of those metrics that I think people are watching very closely. It's going to move around a little bit because of supply and demand, but for the most part, that's when we saw things really get stimulated.
I wonder just in this context, when you think about the kind of PCs or tablets that well, let's put tablets aside. The kind of client PCs or notebooks that Intel, for example, is trying to promote in this market, to what extent does the pricing of SSD, the availability of SSD, help or hurt the ability to promote that kind of new form factor?
If you're speaking of their Ultrabook initiative.
Yeah
I think I'll back away from that for a second and come back to it, if you look at what Apple has done with their MacBook Air and some of their form factors and only using SSDs, it certainly, I think, created a market awareness and a performance expectation that Ed was mentioning that he's so fond of. I think that that's put a lot of pressure on the competing OEMs and the form factors they're trying to bring to market and having that kind of experience, whether it's instant on, power savings, or the raw performance you get from using an SSD. I think there are price points that Intel has in their mind that they're trying to drive in the marketplace. Depending on what capacity you're going to put in there, you can achieve those.
I think that's one of the areas that they're really grappling with or wrestling with.
Okay.
How to get it into that sweet spot to really move, from a velocity point of view, that platform.
Okay, just one last question in two parts. If we were to fast-forward, let's call it two or three years from now, what proportion of your NAND revenues do you think will come from SSD versus what it is today? Secondly, what do you think will be sort of the average SSD density then versus what it is today?
Wow. Well, maybe I'll take a first stab at it and give Allen some time to see if he can chime in with some clarifying comments. I think in terms of the portion of our shipments that are going to SSDs, again, if you look back at the pie chart that Allen had shown, it's already quite high. We don't really have a goal or a target on that. It's already quite high, and we think that it will be at least that high for the foreseeable future, maybe higher. Probably a larger mix shifting to a higher percentage of our own SSDs, of course, as opposed to some of the third parties. With regards to the average SSD density, that's a really tough question to address.
We cited a couple of examples here that run the whole spectrum from our very cost-competitive, low-density M500 drive, that can be, 128, 256 kind of gigs, all the way up to some of the density Ed had mentioned, whether it's in our SATA form factor or our PCIe form factor, one gig, two gig are here today. We will be deploying products with capacities up to 16 in the not so distant future. It really is a function of how those segments play out from a demand perspective. The density question's a little harder to address. Maybe one other thing, just to bridge this back to your prior question, is that we talked a lot about the price and the price elasticity. I can't emphasize enough that one thing that we're seeing that's very different this year than last year, with this very same issue.
In fact, if you go back to the Flash Memory Summit last year, when I had given the keynote, I had a graph in there that had talked about Allen's $1 per gig price point. Last summer, we were astonished that we had come pretty close to that price point, and we hadn't seen the uptick. Well, no sooner did I say that than, as Allen mentioned, the demand turned on, and we really realized that there was a sensitivity there. The rest is history. I'll tell you this year, the thing that we see that's very different is that there is a very large portion of the SSD segments now that is incredibly price inelastic.
Certainly, the enterprise customers are consuming a lot more now, and their value is more around quality, reliability, stability of the supply line, the ability to scale with their business, and provide the NAND capacity that's required to source that. Winning designs there is rarely about price, to be honest with you. The same is true in this explosive data center category, or what some might call the Web 2.0s. The industry has assigned a three-letter acronym to a behavior here that's called large volume orders. People that are building out these data centers really would like to make those decisions fast. They need a ton of capacity very quickly. That manufacturer, and more often than not, it has to by definition, be a NAND manufacturer, has to be positioned to react to that sudden uptick.
There again, the winner there is usually the person that has the capacity, the ability to supply, not so much the one that's hitting the right price point.
Very helpful. Thank you very much.
Our next question comes from Craig Ellis with B. Riley. Please go ahead with your question.
Thanks for taking the question, and nice presentation, guys. Just going back to the slide that showed the range of products across interfaces and thinking about the two markets that you serve and the sub-markets, consumer and enterprise. Can you talk a little bit about controller development and how you approach that issue since it's so critical to performance? Where does it make sense for Micron to do its own internal development? Where do you partner up, and are there opportunities where you just go to third party exclusively?
Yeah, that's a very good question. It's certainly something that I would say over the past several months, we've wrestled with as well. I can tell you that on the enterprise high-performance market that utilizes PCIe, from the get-go, we made the decision to use an internal controller. We've been very happy with that and the progress that we've made with that controller and the solution. I can tell you, going forward, on our SAS drives, we will be using our own internal controller for those products. As we roll out the next version of our SAS drive, it will have an internally developed Micron controller. It is more than a single instance of a controller.
It's a platform that allows us to have a common firmware base that will allow us to have different transport layers that will give us the ability to move rapidly in terms of building on an investment and a foundation of a platform. The real fundamental question is, what's the ability to move that platform downward? Because it's going to start out at the top, where, I think as Glen articulated, the investment's there to support it, and the business is there to support it. We are actually in the process of dealing with how far down on
those controllers and where it makes sense. I can tell you that even if it makes perfect sense, your ability to do that and move those controllers as quickly as the market wants to move in certain instances, despite you wanting to do it, may be problematic. I would say, long answer, but I would say our strategy going forward, certainly on the bottom tier or two, is to have a dual-pronged approach where we use internal controllers where it makes sense, and we use external controllers where it makes sense.
That's great. Thanks, guys.
You're welcome.
Our next question comes from Vijay Rakesh with Sterne Agee. Please go ahead with your question.
Yeah. Hi, guys. I know you, on this PCI, you announced a 16-nanometer ramp. When you look at the SSDs, where are you in terms of the 24-nanometer/19-nanometer/16-nanometer mix that you're seeing?
Your question is specifically within SSDs, Vijay?
Yeah.
Yeah. It's, again, a very good question. What's interesting, I think, as Glen talked about, the different segments have very different cycle times. Down at the bottom of the segment portfolio, when you're talking about cloud Web 2.0 or personal storage, our ability to introduce newer lithos into those markets that move much faster and are more on a cadence with the lithos that we introduce, is much easier. 20-nanometer in that M500, we ramped that very quickly after we brought the technology up. On the true enterprise-level drives, I think as Glen articulated earlier, the cycle time just to get those drives qualified at enterprise OEMs is nine months to a year.
Now you have this almost year lead time to when you get your technology up, your drive completed, the year qualification, and then, by the way, the customers say, "Eh, don't change anything for two years." You've got this delay factor, which leads to the stickiness of the drive in that particular application, and hence, a very different business model that we have to make sure that our factories and our supply chain are able to handle. As much as I'd love to sit here and say, "Hey, that enterprise drive is using our 16-nanometer silicon that we just announced last week," there is that delay factor that takes quite a bit of time. I think the secret for us is to figure out how we work with the right OEMs to minimize that.
I think, as Glen said, it's less about offering them a lower-cost drive. It's more about having the drive there that they qualified the previous year. We balance our portfolio based off of those differing dynamics by segment.
Got it.
This is Ivan. I'll give one quick example, because I think it's perfect to what Ed's describing, just talking to some of our sales guys just the other day about a particular customer who had a chance to move to the next platform, the 20-nanometer platform. Actually, to be simplistic, paid us not to do it. Basically, I will pay you the difference to make margin better for Micron to stay on the 25-nanometer platform as opposed to going through the rigors of the qualification at 20. Again, we make net better margin per wafer on that transaction now. I think the summary is we want to maximize our margins as a company.
Got it. Last question here. When you look at the controller side, obviously on the enterprise side, looks like you're doing a lot of the controllers in-house. As you shrink, obviously the controller complexity goes up, the endurance and reliability requirements go up. Can you talk about how that is helping you in terms of getting more $ on the sale on an enterprise drive?
Yeah, absolutely. I think as Glen showed on one of his slides, it's fundamentally this concept of vertical integration. It starts with that foundational knowledge base of what the NAND technology is really capable of doing. I can say that, and we can talk to customers that we're selling the NAND to explain that, and we do in enormous amounts of detail. With that, there's nothing like having that first-hand knowledge and 25 years of experience of dealing with non-volatile memories when you're sitting down to design a controller.
By the way, you got to remember that when you're designing a controller for the enterprise, the cycle time to design that controller is such that not only do you need to have the knowledge of the NAND technology that you're shipping today, you better know what that NAND technology is going to look like 3 years from now, because that's about the cycle time of a good enterprise controller. You better know what that is. Otherwise, you're going to design a controller that's not going to be in line with where the technology's going. That concept of vertical integration where you understand the fundamental technology, the timeline of where the technology's heading, coupling that with your own internal controllers, it almost guarantees, if you do it right, that you're going to do a better job than anybody that doesn't have that knowledge.
I can tell you that that's a critical piece of our strategic planning process that I think will ultimately enable us to be one of the winners long term in this market.
Got it. Thanks.
I'm not showing any other questions. I'd like to turn it back over for closing comments at this time.
Yeah. Thanks again for all the great questions. I would like to thank everyone for participating on the call today. If you'll please bear with me, I need to read our safe harbor protection language real quick. During the course of this call, we may have made forward-looking statements regarding the company and the industry. These particular forward-looking statements and other statements that may have been made on 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. This concludes today's today's