 Live from New York, it's theCUBE. Covering Micron Summit 2017, brought to you by Micron. Welcome to New York City, everybody. This is theCUBE, the worldwide leader in live tech coverage. And this is our continuing coverage of Micron. Really eliminating many of the overheads that not only were associated with spinning disk but all the protocols and overheads associated with decades and decades of infrastructure that's been built up. And so we've been looking forward to this day for nearly a decade. Mark Durkin is here, he's the CEO of Micron and he's joined by Darren Thomas, who's the head of the storage business unit. Gentlemen, thanks for making our prediction come real. Great to have you today. Thank you, it's a pleasure to be here. And congratulations on what's been, I know a lot of hard work. Mark, let me start with you. Well, let's get right off to the top. I mean, you've been at Micron for the better part of three decades. It was announced recently that you're going to step aside as CEO shortly, turn over the reins to Sanjay. So congratulations on a great career. You told us off camera, you're not going anywhere. But day-to-day operations, you're turning over to Sanjay. Take us through why now and some of your thoughts and reflections on that. Well, first of all, it's been a wonderful, fun, exciting, adventure-pack ride and I've enjoyed every minute of it and every minute of leading this great team here at Micron. The future for me is really for now just to spend a little time away from the job and focus on my family, but I'm sure there's going to be lots of new adventures in the future. The timing is really just right. Micron is firing on all cylinders at the moment. When I think about where we are from a technology perspective and the great progress we're making there on lots of new memories as well as our major incumbent memory technologies, it's amazing to see the progress and the momentum. The products and the product portfolio is qualifying and has huge interest with all our customers. So we're excited about that. And finally, we're making great progress on the system level solutions that we were talking about here today, not only in the storage market but across all sorts of other markets as well. And you put all that together with what is a really strong tailwind from the markets themselves and a very solid set of circumstances in the marketplace. And it's just a great time to make a transition like this. And so that's what we're doing. And Sanjay obviously has background from one of the companies that you acquired. He found it. Tell us a little bit about him and why. Yeah, Sanjay is uniquely qualified I think among everybody in the world to run a company like Micron. As you mentioned, he's the founder of a great memory company himself which he grew to really a wonderful systems memory company. That's obviously where Micron's going. Our future is about system solutions for memory, memory subsystems and systems. And he's got a strong technology background. He understands manufacturing and he understands what makes memory successful and its importance. Our end markets, our customers. But most importantly, he's got a great passion for this business. And Sanjay is a guy that's had great success. He doesn't need to do another job or run another memory company. But he's excited about the opportunity and coming to Micron to work with the great team and the resources and the technology we have. And so you put all that together and it's just a phenomenal solution for Micron. And another reason why now's the right time. Excellent. Now Darren, of course we've known each other for a while. You were one of the early, you know, Cube guests when you were at Dell. But things have changed a lot since then. Well, the first question is, how many Ts are there in Zettabyte? I think it's two, right? Three, yeah, okay. So you come to Micron from Dell. Obviously, lots has changed since you joined. But take us back to the sort of strategy of this whole notion of enterprise storage. It's a small, relatively small contributor to Micron's overall business today, but it's huge potential in the future. Yeah, I think the strategy is fairly simple. Customers want this. Our customers want to be able to address us directly as a fab shop. They want that technology. They want access to it. But we can't speak to them in fab terms. We have to speak to them in the terms they understand, which means that we have to speak like we did today at this event. And when I was hired, Mark made it very clear that he wanted to build an enduring enterprise business that was based on our ability to connect directly with these customers and deliver solutions to them as they need. I mean, make no mistake about it. We still sell the components, but as time has gone on, we sell more of the subsystems and we fully expect to be investing in the solutions through our partners and with our partners. We're not trying to circumvent anybody, but in this particular case, the product you saw today, it takes the insight of a full micron from our deepest technology people in the NAND and DRAM all the way through the subsystems people that understand all the abstraction layers that are in there, all the way through our software teams. When I came to micron, I was living in Austin. We built an Austin facility that has a software team there that understands OpenStack, VMware, Microsoft, Linux. They are experts at that. They come from that industry and that team is busily helping us construct the processes that allow us to build a solution that has all this acceleration and that you've never seen before. Well, so you sort of referenced, hey, we're not trying to end-run anybody, but there's a natural progression in the industry. David, we've talked about this for years when we first saw the Fusion IO technology and then the acquisitions you made. You said, the industry's gonna change and it's gonna collapse. But you still probably had a convinced mark that it's okay that what we're doing there. What was that discussion about? Well, it's not just okay. It's imperative. This is the future of memory. Micron's got a rich history and culture around being a strong technology company, technology push company and a strong manufacturer. But the reality is the future of memory is about customer intimacy, customer value through really differentiated solutions that address the workloads and the problems that the customers are trying to solve across a broad, diversified set of end market segments. And so we have to do this not only in enterprise storage but we have to do this in all the end markets we serve in order to deliver the value that the customers want from the memory solutions. And I couldn't be proud of the job, the storage team's done here, but we've got great progress going on across the mobile solutions with the mobile business unit and computing and networking and automotive to support ADAS systems. And wherever you look, there's memory today and wherever you look, there are customers that want more from their memory and solutions are the answer to that problem. So I have a bad habit of asking multi-part questions and I want to ask you both sort of where we've come from, sort of where we are, where we're going and sort of the what's changed question and then David, I want you to jump in as well. But I'll start with Mark. As you mentioned, there's real tailwinds in your business right now. Chip prices are up. There's been consolidation. You guys talked about that today. Moore's law is attenuating. I mean people maybe debate that, but I think it's pretty clear. All good things for you guys. So business is very strong, stock prices up, investors are sort of starting to get the story. There's always a concern that it's a cyclical thing and prices are gonna fall again. So where are we and where are we going? What's different now than sort of 15, 20 years ago when you were just a pure PCDRAM company? Well, first of all, I'm not here to say the semiconductor market is no longer cyclical, but I will say that the memory market is gonna be much less cyclical than it has been historically. And the reason for that is all the things you just mentioned. It's a slowing Moore's law, which mitigates the demand balancing problem that we have when we create supply automatically by staying at the efficient operating frontier or by developing new technologies to enable new answers. The end markets have diversified and so we're servicing so many more different types of customers within a fixed supply bucket. And the memory suppliers have consolidated so there are less of them trying to balance the equation and the combination of all those things really makes for a more stable marketplace. Before anybody starts trying to think about how do we now differentiate our end products and the natural tendency as Moore's law slows to have more differentiated solutions, more value added solutions for end markets because we're not competing with something that's just 30% cheaper and 30% smaller next year. We're really getting the opportunity to work with the customer and solve their problems for them. Okay, now Darren, when you were at Dell, Flash was barely talked about. I think the first time we did a CUBE interview, we were talking a whole different set of issues. And David, you were one of the early ones on that. So maybe you can take us through sort of what you've seen as the change in the last several years. Yeah, well, when Flash first came out, it was small and expensive. And the industry that I come from likes big and cheap. And so small and expensive has a place in it, but everybody saw the potential of it. Everybody could see that this technology was just shattering the performance barriers and power and cooling and battery life and all those things had tremendous value. So they kind of tolerated that. Now, fast forward to where we are now, we are big and we are close to, in many applications, less expensive because it takes a lot less of our drives to get to any certain performance number. So we have become the big and inexpensive device and that has just opened all of the doors. I mean, one of the reasons in the storage market that the market is such a tailwind is because we have the client people who absolutely understand the battery life value and the small form factor, but you out now have the web scale guys that understand the cost and performance benefit value. And you have the enterprise companies that are wanting to solve the performance numbers that we saw today in our presentation and delivering a solution that's four times faster than anyone's seen is a major breakthrough, but it might have been lost that we can do it at a lower price. And so this is the perfect storm for storage. We have a technology that is just better on every dimension, better size, better power, better cost over time and obviously better performance. And so this perfect storm is what's happened now and we weren't quite there and we weren't quite there when I came to Micron, but you could see, I mean, when I walked in the door, we had this 3D NAND technology well understood and it was just a matter of getting it to the market and that product is now fully invested in our company. It's in the market and it's making these changes occur daily. So a question I have is on the systems side of things because historically the companies that have brought systems have been the IBMs, the HPEs, the other members of that community, but you're bringing in today a systems solution which is going to affect the whole architecture and the whole stack from top to bottom basically. Why Micron as an innovator in that area? Why, what's different? I mean, you're in a very disruptive point in time so that's a good time to come in, but do you think you've got the pedigree and the capability of pushing that? Yes, yes, today Micron is a much more nuanced and sophisticated company than we used to be. We want to keep that pedigree as a lean mean street fighter that develops great technologies and is brutally efficient in manufacturing, but our vision and our internal culture has evolved. We see the value in delivering differentiated solutions these end markets and it's not just by choice, it's really imperative. It's what the customers need us to do in order to take advantage of these technologies. If we don't do it, either the solution will not exist or a competitor will do it. And so really our job is to be really closely partnered with these customers across a broad swath of applications and help them get their solutions to marketplace. And that's what we're trying to do and I can't tell you how proud I am of the team that we have built and the capabilities we have in-house today to help affect these changes and deliver value for our customers. In the time we have, let's talk a little bit about NVME and NVME over fabric, it's kind of why we're here. So during the timing of that, it's been kind of a shipping in Q5 concept for a while now and we're here. You talked to Justin Stahlmeyer, I talked to him last night at the Yankees game and he's running in his lab, he showed me some of his test results, he gets daily updates from his team and this is a real deal. So talk about the timing and the impact on the business. Yeah, the timing of this particular technology and capability really needed NVME and NVME over fabric protocols to evolve to the level they were enterprise worthy. And NVME, you can imagine, any protocol that comes into the storage world, most of them have taken decades to perfect. And I mean, we're still using scuzzy protocols from decades ago and even fiber channel uses scuzzy protocols. So these are all very complex technologies that have built in it, that error-free operation and recovery systems, self-health checks, there's a lot of software in there. And NVME has really come of age, if you will, in these latest releases. And that's what you're seeing is prior to this, we could have built an NVME solution but it wouldn't have been enterprise worthy. And what you're seeing now is we can make one that the enterprise customers will not only love the performance of but embrace the quality and reliability of. And so that's what's really happened right now. And to echo what Mark said, I don't think anybody else could have done this because you had to understand how the fabrics worked. You had to understand inside how the NAND works and how the DRAM works and how to optimize those things. So what we discovered was we partner with all of our normal OEMs, we partner with all those guys, but we have a much deeper insight into this. And because of that, we can move much faster, we can make decisions, how we designed and built 3D NAND, how we're doing our NVDEMs, all of that was built understanding we would get to this point. And that was what was required to be here. So the key thing to me is the NVME over fabric, the NVMEF, which is all the worst acronyms. Yeah. So that seemed to me the absolute key. For the first time, and Fusion IO tried to do this and David Flynn's old company, they tried to do that, but they couldn't do it across a fabric, across all of the computers, $250,000, that wasn't possible in that day. So what is it that you've introduced in that fabric, that architecture that enables you to bring to fruition what people have been trying to do for the last couple of decades? Well, again, it's the maturity. We partnered with Melanox, and Melanox has had enough time to really get the protocol working well inside their silicon. And believe it or not, you have to be ready for prime time. The enterprise will not tolerate a science project type technology. And prior to this, that's kind of what it was. Everybody knew it was coming, but it really wasn't ready. And I think what we've seen here is probably since about November, this past year, we've seen this technology mature to the point where we built it in our labs and we could run it very efficiently. And we've been tuning it and debugging pieces of it. And just so you know, we're in beta phase right now. So all those customers you saw that have this in their lab, we fully expect them to find issues that we will have to continue to work. And this product will be ready for general consumption later this summer, so. So we gotta leave it there, but Mark, I want to close with you. I feel like, you know, you've been here a long time. I feel like you're going out on top. A little bit of like the San Francisco 49ers, Bill Walsh, handing the reins over. The team's in pretty good shape, right? You got a new QB coming in. But your last, their final thoughts, just some reflections. And especially in terms of the latter part of that, like I say, you're leaving it in good hands and there's a lot of interesting things happening to market, AI, autonomous vehicles, IOT, mega data centers, so your thoughts. Well, I'm just, I'm so excited for the future of memory and the memory industry, and in particular for Micron and the Micron team. You know, Darren's a humble guy, but getting solutions like this to market first. Yes, it takes a lot of hard work. It takes engineering excellence. It also takes vision and leadership. And what they've done here is just one example of a lot of great stuff that's going on at Micron. So I couldn't be more excited about where we're going. I couldn't be more excited about the role memory will play in the future of computing and things like cognitive computing in particular, artificial intelligence, and really changing the world in very significant ways. Great, well gentlemen, thanks very much for coming on theCUBE and a great event and congratulations. Thank you so much. Thanks for having us. You're welcome. All right, keep it right there, everybody. We'll be back. We're talking to practitioners. We've got a CIO discussion coming up when we're going to dig deep into the technology. So leave it right there. This is theCUBE. We're right back. We're at Micron Summit from New York City.