 Okay everybody, we're back. This is Dave Vellante of Wikibon.org. This is siliconangle.com's continuous coverage of EMC World, we're live in Las Vegas. This is the third year for theCUBE at EMC World. We started here in 2010, we were in Boston. EMC's been great, great supporter of theCUBE and just keeps getting bigger and bigger every year. This year EMC World, it's I think over 13,000 people. We were just watching the keynotes, Joe Tucci, Pat Gelsinger, a lot of energy and we're going to talk flash. We're here with Mark Sorensen, who is the Senior Vice President of the Flash Products Division at EMC and also joined by Barry Ader, who runs marketing for that division. Gentlemen, welcome. Hey Dave, good to see you again as always. Yeah, thanks so much. Last time we were on theCUBE in the Wikibon headquarters. That's right. That was right around the time of the VF Cash launch. We covered that pretty extensively, you guys. I think, really, I think legitimized that whole concept, a large player, now putting a lot of resources behind it. You had a lot of startups doing a lot of really cool things. Yep. But you guys brought your largest to the marketplace. Give us an update, Mark. What's the customer reaction? So, let's see Dave, when I think when we last talked with you at your Marlboro headquarters, we were just announcing VF Cash, right? This is PCI-E flash card in the server, controlled by intelligence software to give you the performance plus the intelligence and protection that customers are looking for. We also more recently announced the Project Thunder. Thunder is a flash appliance for low latency, high performance applications, and that goes to beta, this quarter, this month, actually. And then, of course, Todd, not to talk about flash, EMC's most recent acquisition, Extreme IO. Yeah, so, a lot of pieces. You started with Enterprise Flash Drives. You announced VF Cash, PCI-E card, read-only, Thunder, starts to share stuff. Extreme IO, where's that fit? Well, let's actually take it, if we could take a little bit of a step back to, I don't know, Barry, was it 2005, 2006, when we first started building flash into our storage arrays? 2005 is when we first started looking at it. 2008 is when we first actually were able to put flash drives into a symmetric, followed that up real quickly with Clarion and Fast. One of the things we see with flash is we see it as something, as an enabling technology that should go into a lot of different places, whether it's in the storage array, as a hybrid array, and that's a new term we're starting to use, hybrid array, an array that allows you to have both flash and hard drive technology. But we also see flash in a number of different technologies, like all flash arrays, and that's really where Extreme IO comes into play. And then lastly, or actually another couple of things, is in the server network, either as Project Thunder or as VF-Cache. Many different technologies from many different places, and it's really that enabling technology that allows you to use it in so many different ways. Yeah, so you've seen that flash hierarchy start to get more granular. Exactly. There's kind of a big discussion going on in the community around, all right, should it be a storage-centric architecture? Should it be a server-centric architecture? Is the answer yes? You know, the answer is yes. Indeed, that flash, as I've said to you many times, we think it's going to be a ubiquitous technology that will find its place across a number of different deployment options, whether it's in the server, whether it's in the network, whether it's in a hybrid storage array, or whether it's in an all-flash storage array. If you actually look down below, we actually use flash more surgically than what the press and pundits talk about. We use flash as metadata repositories to speed up file access. Our backup products use them to place indices to speed up deduplication. So it actually goes down another level on that, but of course, these big acquisitions with big price tags, you know, people get excited about it, and they want to pit one versus the other. The answer is it's all of the above. You'll see EMC introduce or continue to deploy its intelligent software called FAS that provides this umbrella across all of these platforms and technologies. Yeah, I mean, you said something, they want to pit one against the other. People didn't pit SATA drives against FC drives, right? I mean, so, you know, they have different roles, different cost structures. Is it the same here? It's exactly the same here. So it's how do you figure out there are lots of different technologies? How do you figure out where to place it at the right time, at the right place? And that's where fully automated storage tiering comes in. So there are a lot of different technologies, a lot of different architectures, a lot of different places you can put flash, but the question is, how do you decide? Let the system decide by leveraging software to move the hot data to the right place at the right time. So that was my follow-up question. So it seems to me, it's, if you've got this elongating hierarchy and all these functions moving or all that, it seems like software is the glue that holds it all together. I think I'm hearing fast is that software. Is that correct? Or is there a white space there that needs to be filled? I think fast is the fundamental umbrella technology that helps move data across. Clearly management plays a role. You see us do things around ensuring improved data endurance. Those are other concepts. But certainly fast is probably unique to EMC and something that we think is a key technology to provide both performance as well as cost efficiency. Because you want to use the right tool for the right job. Flash is, let's face it, it's still an expensive technology when compared to disk. And for the foreseeable future, we believe that disk will always have a cost advantage over flash. So if you have data that is doesn't have the performances of flash, you need to put it on spinning disk. And that's why we believe that hybrid storage arrays will still be probably the predominant deployment model above flash and disk across for information management systems. Yeah, Pat Kelsinger said it's a thousand times faster, but a hundred times more expensive. So obviously you need software to put the right data with the right device characteristics. You guys have seen a lot of technology disruptions over the years. We saw the client server. Joe was talking this morning about waves. He gave us a great history lesson. Yeah, that's great. I really enjoyed that. And the reality is that there's a lot of money to be made to those who can identify, to spot those trends and take advantage of them. Mark, talk about flash as a disruptive opportunity. And at what point did you guys decide that you were going to be all in on flash? And how will it change your business? Well, I think, you know, if you look back, what's the disruption of technology? It's really, you know, the technologies that'll enable cloud computing. Providing an obstruction layer across your physical devices enabled primarily by virtualization. And we believe VMware is that the best virtualization technology in the industry. And so that's kind of the foundation. Now as you do that, you create several problems. And those problems include some IO bottlenecks that are occurring because of virtualized environments, plus exacerbated by the continued performance leaves that our microprocessor partners make there. So it's a disruptive technology, but it fits in a larger picture, a larger ecosystem. And that's in a cloud environment which you can define as an abstraction of computing environment driven by virtualization. So, yeah. And I wanted to follow up with that. So that, your answer was very good. And it was sort of, I wanted to ask you from a marketing challenge. You guys are, the company sells a lot of disk. And here comes this flash disruption. You're the marketing guy for this division. How do you balance that sort of, you know, message? It's all about use cases back for the customer. So if you can have different tools in your bag. So if I'm an EMC sales rep, I'm loving life right now, because I have the different tools in my bag to be able to solve one problem and solve another, you know, no matter what the customer's trying to achieve, they have the different tools to do that. So it's actually, from a marketing perspective, actually pretty fun. Because now you get to go into, you know, new markets, you get to go against new competitors, and you get to leverage all the stuff that you've been bringing to the game for many, many years. It's kind of actually building on top of each other. So it's VMAX plus VF cash. It's VNX plus VF cash. It's not one or the other. What's amazing to me is, you throw here statistics like only 5% of the world's data is candidates to flash. And maybe that'll grow over time. But still, it's relatively small. But the use cases are infinite. I mean, it used to be, okay, we're going to, we're short stroking FC drives. And so that's the use case. But Mark, you mentioned, you know, indices, metadata. What are the customers telling you guys? Well, the customers basically want us to decide where to use Flash. So when we first introduced Flash technologies in our storage way, we gave it to them and said, here's some fast storage device. Go put your hot data on it. Excuse me. And the issue was, A, they couldn't figure out what their hot data was, or B, that changed. So what was hot one day was cold the next. And so what really accelerated, and you can tell by just looking at our revenue trends and our capacity shipments of the leverage and use of Flash was the introduction of fast. Intelligent software that decided real time, what was hot data, put it on Flash, what was cold data or cooling data, we'll get back to spinning disk. So we get back to that intelligent software that'll let you use right tool at the right place, right performance, right cost. And to add to that, so fast is also take fast to another level and think of it as, well, how do I leverage this for caching devices? So cache is interesting, right? Because I still get to use my back end, but I get to use and figure out the hot data as a cache. Now, PCIe technology, great Flash technology, been around for a couple of years, but again, back to what Mark said, customers had to figure out, what data should I put on that PCIe card? Now with VF cache, they don't have to figure it out. The system can figure out for them by determining what's hot, what's hot today versus what's hot tomorrow and still get the protection of their back end storage. You don't have to move the data, they can just intelligently have the hot data at the right place. So that's interesting because there was a period of time where I think customers were afraid to give up that control. What I'm clearly hearing is that's changed. They are demanding that you take care of it. Yeah, I think they have, and certainly as we've shipped a large amount of fast software, it's proven itself, it's very reliable, there is some control by customers on what gets moved, what are the policies, how much gets moved, but today it's been a very, very successful launch of that technology. Dave, I want to go back to one point you talked about, you're kind of half-kitting Barry about the marketing. The thing, the challenges that Barry has, I think, is that if you look at EMC's portfolio, it's very, very broad. Most of our competitors are one-trick ponies. For them, it's off-lasher A, it's going to conquer the world, nothing else. Or PCIe flash in a server, no other room. Or a box in the network, this is the way you deal with this net. We're not religious about this, right? We recognize that you have different needs for different applications, different use cases, and again, with intelligent software to bind them together, we provide the flexibility for customers to really deliver the things that they need in the right place, right location, right time, right cost. Well, that brings up an interesting point, because we had Pat Gelsinger on last year, he said, you know, we fell a little bit behind in Flash, which it was candid, and we were impressed that he said that. But in thinking about it, think about what you just said, yeah, maybe you fell behind some of the startups, but I don't think you're behind the larger players. No, it's actually a very interesting point. Yeah, it's actually a very interesting point. To Mark's point, so your big established players you would expect are going to have big portfolios. I mean, IBM's got a big portfolio, HB, they can buy companies, they're not there yet in Flash, so you guys are ahead of the game relative to the whales. You're buying companies, you know, tuck-ins, big tuck-ins, they, hey, it gets bigger. By the way, I've said acquisitions oftentimes are much better used in cash than R&D, because R&D's really risky. And there's a big land grab here now. So, that, I would expect that that portfolio of competition is going to start to build out. You know, what are you seeing there as far as the competition? It's really today, these startups, some of them haven't even got products to market, but they're doing a good job marketing, some of them are reselling products, like for instance, from Fusion I.O. What's the competitive landscape look like? There's really a lot of customer confusion out there. So, it's interesting that you said the comment about the whales earlier, because if you look at the Flash landscape, you know, and we look, we have some, you know, material obviously about the competition, where do you find some of these other bigger companies? They're not out there. They're not selling a lot of Flash, so the Flash landscape right now are a lot of the startups out there that are trying different things that, you know, they may not have a GA product yet, but that's really where the Flash landscape is. There's a couple of other bigger guys, like you mentioned, like Fusion I.O., you know, that have been around for a little while, you know, that are stuck, you know, just simply inside of the server, but some of the whales know where to be found. Yeah, and they're talking, I mean, conceptually, you know, Flash, you know, moving to a closer to the server, that should favor, you would think, a lot of the server guys, but from a time-to-market standpoint, I think it's fair to say EMC's clearly beating them, and why is that? You guys just like Andy Grove, paranoid, you're watching the startup saying, hey, we fell behind, we got to go, go, go, go, go. Well, I'll tell you the story of how it occurred. I mean, we were feeling very, very good about ourselves after shipping umpteen petabytes of Flash in our storage arrays, and then, you know, one day, Pat Kelsinger came to me and said, Mark, I want you to form a Flash business unit. I said, a Flash business unit, Flash is, yeah, it's a next-generation disk drive. Is that a business? And he said, yes. And we sat down and we talked and looked at, you know, kind of how Flash was going beyond just being a disk drive, but a transformative technology that was going to be in the network, going to be in the server, was going to have mission-critical intelligent software, whether it's VF Cash, caching technology, or Fast, it was bigger than that. And so, you know, you got to give some credit to Mr. Kelsinger. And of course, he's very smart when it comes to microprocessors. So we understood the transformative and the disruptive change of that. And yeah, we're a little bit paranoid. Some of this stuff, we're building ourselves. VF Cash, we built ourselves. Thunder. Thunder, we're building ourselves. We talked about Extreme A.L. We saw this, what we think is the potential jewel in the marketplace, and we just grabbed that. Fortunately, our business is good enough that we have enough money in the bank to go off and do these type of things. So I, you know, I think there is a little bit of paranoid. I think there's a sense of urgency. And, you know, I think we got into Flash early and saw what it could do. So the role of your, it's a fascinating move that Pat made to put you in charge of it because you're a seasoned operational guy. You've had a lot of credibility and a lot of different divisions and a lot of different companies over the years. So you've got street cred. So they put somebody like you in charge that can organize a group and move fast. What's the role of the division? Is it as an accelerant or more than that? Well, you know, certainly ownership and development of technology is driving strategy across the corporation because, you know, we call it the Flash business unit, but everybody owns a piece of it, right? So we saw some of the stuff that Brian Gallagher's doing in terms of integrating VF cash. You guys are talking about fast. Symmetrics fast crosses our storage arrays, the use of this technology. Partnering, right? Partnering with the Flash providers is very, very important. So I think it's an accelerating, I think it's strategy and of course some actual development as well. And of course, Barry leads the effort in terms of our messaging and product management of those capabilities. Yeah, and it's actually been a lot of fun. So we've been in this business unit for a little over a year. Think of it like a little bit of a startup within the company. You know, the startup part of it is the fact that we can bring new products to market, but there's also this advisory part of it, the strategy part of it, that we still do work with all the other divisions as from an overall Flash strategy, a lot of fun. It's an interesting dynamic to me because EMC is a lot of alpha personalities. And you know, I mean, the competitive externally, but I imagine there's a lot of competitiveness internally as well. But it seems like you guys, I guess by the very nature of what you're doing, have to partner with others internally. We do, and you know, for some of us who've been here for a little while, I mean there are colleagues, but frankly, most of them are friends as well. And we battled out a little bit at the end of the day. We do the right thing for the company, our customers. And I think it's shown the progress been pretty good so far. So if we fast forward a couple of EMC worlds from now, what do you think this world's going to look like? You know, today you've got a disk base that's clearly in transition. You've got Flash popping in everywhere, but it's not holistic yet. Are we going to see that over the next couple of years? What do you guys think? I think it'll definitely get a lot more holistic than it is today. We can see the path to where the movement of data throughout the entire ecosystem and infrastructure can be there and can happen. We'll also obviously see new technologies, whether it's phase change, memory is the next big thing. So we'll see new things coming up. Yeah, you don't care, right? And it's real fun, it's real exciting. Yeah, I mean, from your standpoint, you don't care what the underlying technology is, as long as it's the lowest cost and highest performance. As long as you can end up delivering something back to the customer to meet their business goals. If I look a couple of years out there, certainly what Barry says is true, the underlying technologies will change, but I think the biggest thing that will happen is cloud will move from to some degree kind of a concept to really an entity. And so we'll be talking about fast hearing into a cloud, whether that cloud is from a cable company or a telecommunication company or something. I think you're going to see cloud be much more developed as an entity versus what today it's, it was still conceptual for a lot of people. People are moving to it, they're taking the steps around virtualization to get there, but it's still fairly new and still concept, it's still one of those kind of waves on the slide that Joe showed. I think you'll see much more reality behind the cloud and I think you'll see Flash playing a big enabling part of that. Yeah, and you're obviously seeing a lot of press too with the valuations I mentioned before. You've got companies like SolidFire that get hundreds of millions of dollars in valuation, having shipped a product yet. I love the SolidFire messaging, I love their vision, but haven't shipped a product yet. Right. And so, extreme IO, very early into the marketplace. You saw that with IO Turbine. You see violin memories about to do an IPO with an $800 billion valuation. So you guys, you must sit back and kind of chuckle about this stuff because you've seen it before. You guys are largely driving the market with these types of transitions. So, you know, you're in the cat bird seat, you don't have to comment. I know you're in trouble, but it's very interesting dynamic. I personally, I think that it's great. I think people should strike while it's hot, but I don't think it's sustainable. I think that companies like you are going to really absorb a lot of these innovations in vent. You're going to tuck in, you're going to buy, and things will settle down eventually. You know, we work in a very interesting industry, if you will. I was looking at those waves that Joe showed, and I wasn't in on the mainframe wave. I did come in kind of towards the end of the mini-computer wave with Digital Equipment Corporation. And we've seen a lot of change. We've seen a lot of change in all this time. And I think, I don't see it subsiding at all. I think we're going to continue. And frankly, we're going to be in this game to make that happen. Are you sure you weren't in the mainframe? I was there. I'm pretty sure. I saw you. I was there. I saw you. You know, you was definitely there. I had heard about mainframes, but I'd never touched. I'd never logged into a mainframe. Oh, boy, what's that? Fantastic. All right, gentlemen. We're talking about new changes in I.O. architecture is driven by Flash. Mark Sorenson, bariator, thanks very much for coming to the kid. Thank you, Dave. It's always a pleasure to see you guys. Thank you very much. We love what you're doing. And keep it right there. Everybody will be right back after this word.