 Live from the FIIA Barcelona Grand Via Compensator in Barcelona, Spain, it's The Cube at HP Discover Barcelona 2014. Brought to you by headline sponsor HP. Here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We're bright eyed and ready to go for kind of the second half of day, too, here in The Cube at HP Barcelona. This is The Cube. I'm John Furrier, my co-host Dave Vellante, where we're here to extract the synth from the noise HP Barcelona in Spain, HP Discover 2014. Our next guest is Cube Vellante, always favorite to have on David Scott, SVP General Manager of HP Storage. And David, welcome back again and just want to say, you know, thanks to all your support with The Cube. This is our fifth year. You got you sponsored by us originally and we've been part of The Cube legacy and we really appreciate it. It's become a cult here at HP and people love to get on and thanks to you and your team. So I appreciate that. But here at the event, let's get back, get it right into it. So storage is again featured on stage, real bright spot in HP. It has been since you, you've been part of your team came in. But last night, you know, you can just see the community of storage and with converge systems. Storage, that culture is starting to radiate out. You're starting to hear people say, innovation, that startup mindset is kind of breeding in the groups. So give us a state of the state of the state of HP storage. How's it spreading? How's the converge stuff? You were doing the remote interview from from stage. What's going on? Give us the quick update of here at HP Discover 2014 in Europe. Yeah, sure, John. I think first of all, a lot of private has to go to Meg personally because she's really focused on innovation and actually investing in R&D to make innovation reality. And there's nothing worse than acquiring a company and then not actually investing in that company to make the most of it. And that's not a mistake that HP has made over the last few years. Certainly, you know, in our entire HP converge storage portfolio, there has been a lot of additional R&D investment. And we've seen the fruits of that coming out this week at Discover in our announcements around converge flash arrays, you know, innovative flat backup schemes to make it really simple to back up your storage devices in our moves in software defined storage and hyper converged appliances to kind of fend off the challenges of the new tannixes and simplivities of the world. And that investment is not just in storage. It's also in converge systems. And we're seeing a really integrated approach that is making the whole idea of converged infrastructure solutions a reality for the new So where do you see the performance and simplicity equation intersect? Because you know, there's all kinds of under the hood stuff that we can get into. I'm sure Dave's going to ask some pointed questions on, but I want to get more to the high level message, which is, you know, converge sounds like a great story, simplify things and create this new DevOps like environment in companies, whether it's enterprise or large scale. But how do you take something complex and make it simpler and still drive performance? Because at the end of the day, the environments are changing out there in the enterprise. Well, I think there are fundamentally two strategies that have been put forward in the industry today. One strategy is to say the software defined data center, all you have to do is introduce a control plane. And that control plane can handle the underlying complexity of different types of storage architecture, etc. And that's, you know, the philosophy of a lot of the major vendors in the industry. It sounds easy. It sounds really easy, right? But we don't believe that hiding complexity is the same as delivering simplicity. And we also believe that the only way to deliver a consistent and coherent data center architecture that's incredibly efficient is to not just have a software managed layer as a control plane, but also create simplicity underneath the covers, eliminate these fragmented silos. And that's why we've built HP Converged Storage on this concept of polymorphic storage architectures, you know, a single architecture for primary systems based storage, a single architecture for software defined storage, a single architecture for data protection services that span across both. And if you deliver simplicity at the bottom layer of the foundation, you get the data center efficiencies that people are looking for in the new style of IT. So does that basically sounds like an operating system? You always want to decouple and make highly cohesive subsystems. Is that what you're essentially saying with that? Yeah, architecturally, at least. I think one of the, I think you're right. I think one of the reasons why people can be resistant to the philosophy we put forward is that they say that isn't it impossible to have a single architecture that can do so many things well? And the answer is it's not impossible to do it if you have the right design approach in the beginning. So people often say to me, look, David, you can't really claim to have Freepar as a flash optimized architecture because it was designed at a time when hard disk drives, you know, were prevalent and flash didn't even exist. That's a true statement at least. And it's a true statement. But what it doesn't recognize is that we didn't buy, build an architecture at Freepar for hard disk drives. We built a storage I O processing engine where we were looking at all of the issues that are associated with I O processing. How do you make sure you paralyze as much as you can? How do you avoid the front end contention from the back end contention that is the bane of most storage systems designs? How do you deliver efficiency without compromising performance? And if you build a storage I O architecture that achieves all of those objectives, fundamentally, it doesn't matter what the backing store capacity is, and that's why we've been able to move from hard disk drives to flash and flash optimized solutions and will be able to move to non-volatile architecture. It's got a bunch of server guys built it. But that's exactly the reason. If storage guys built it, they never would have come up with a storage. The storage guys would have built a hard disk drive based array and we bought a storage I O engine. And by the way, those techniques, I'm pleased to say, also similar techniques that happened through our store virtual software defined storage because it was another innovative company, Left Hand Networks that turned around and designed that architecture and they pioneered software defined storage back in 2007. They had the first virtual storage appliance in the market for VMware, what is now six, seven years ago, with the foresight and the vision of how software defined storage would evolve over time to leverage industry standard servers and be hypervisor agnostic and offer a different model, the ability to run compute applications as well as storage on the same platform. And so now we have a second polymorphic architecture, but in the software defined storage space that again has unique capability to use. You can understand why people were skeptical. You didn't go out and buy an all flash array company. Like some of you competitors that EMC did, IBM did, you did not. But you didn't have the money to buy at the time. Meg said, we're not doing big acquisitions or any acquisitions really until we pay down the debt. So you sort of were forced to do that. So people, I was a skeptic. I said, all right, well, and then people were using your marketing term of bolt on because everybody was bolting on, you know, virtualized storage, then provisioning onto their legacy system. So they were using the marketing judo move against you. Yes. You just smiled and said, okay, stay tuned. And now then you come up with a product and even still there are skeptics. I'm still watching. But today we got a proof point. Leaphead low came on from Sony. You interviewed him yesterday, took your all flash array, popped out a symmetric VMAX. VMAX, yes. Didn't go with Extreme IO or other, not an all HP shop. So that's a proof point. Obviously want to see more, but maybe you could talk about that a little bit. You got to be pleased. Well, ever since we announced 7450 or flash array, particularly with the new FinD duplication software, our sales have absolutely gone through the roof with that technology because we solved three problems that stopped all flash optimized arrays getting into the mainstream. One was the cost point. The second was enterprise scalability and the third was enterprise resilience. That's that. One resilience. And we were able to bring all of that together as an alternative to traditional hybrid storage arrays like EMC VMAX. And now we're making it incredibly easy for people rather than moving from VMAX to VMAX 3, going through being the last people to move to the legacy architectures for traditional data centers. They have an easier move to a three part converge flash array using our online import technology for minimally disruptive data migration. And they can position themselves well for the new style of IT. The interesting thing that we learned from our discussion in June was you used a term called escape velocity. And I'm always fascinated by these waves and the virtualization wave. We saw some great exits. Some guys obviously participated in that along with many others. Compellent and Isilon, data domain in a sense. And you said that you didn't think that the all flash array, the pure place would be able to achieve escape velocity. Yes. Because yourselves and your major competitors all have solutions. Do you still feel that way? Do you feel like your competitors solutions will eventually or are there or will get there? And that'll make it tough for these pure place? You know, my contention was if you're a startup you really need a four year kind of period to reach escape velocity where you have a significant architectural advantage. And for most of these flash array startup companies two years was it. And I say was it. It's passed. We have now caught up, exceeded in critical segment areas the all flash startups with three past all served converge flash arrays and all flash optimized solutions. And once you catch up architecturally with a pure place startup they have nowhere to go. They'll continue to have momentum. They'll have an aggressive sales force who's focused on living or dying putting food on the tables based on whether they can sell. But for the mainstream use cases and for the big market opportunities I think three part now is probably positioned better than any other company. Sorry HP with three part is now positioned better than any other company in the industry to winning the all flash optimized market. With the amount of money that's been this is something you know a lot about with the amount of money that's been raised which dwarfs what you had to do which at the time you said was a massive amount of money and memories. I remember now considering myself cheap at $183 million versus $500 million. Yet the investor, the board and the investors were far south of that. You said guys forget it, it won't work. I'm not interested unless you really are committed to this thing. Now the valuations dwarf what you saw. So the only option is public markets. So we're going to see that presumably it's going to happen soon but you know what that's like. As a small pure play it's very very challenging. Life changes when you go public. It's going to be interesting because you have examples out there of kind of pure play startups that have just gone public, like Nimble as an example. And one of the challenges we're going to see whether they can execute it effectively is big investment that they're putting forward to try and grow the top line as fast as they can. But the bottom line continues to drop dramatically and making that corner turn as a public company is going to be very very painful. And it's going to be interesting to see if newly complete companies like Nimble can do it. Yeah well I mean again you saw this in three part when you were public completely any free cash flow had to go back into building out the channel and building international sales and so forth. And then if a company buys one of your pure play competitors all of a sudden your distribution advantage just went oof because HP's got a massive or IBM or EMC or whomever. So it's going to be really interesting to see so what is the state of the storage business? I mean is it just the rich are going to keep getting richer? Or is there, no I think the dynamic if you remember two years ago at Discover we introduced the three part 7000 series. This mid-range platform but with kind of tier one mission critical availability and resilience. But priced at a point that anybody could afford. And when we introduced it one of the things that we put forward was we thought it would have a significant impact on the high end storage market. If you now look back from two years forward at what has happened over the last couple of years you've seen the high end storage market collapse. You see it in the data that IDC is constantly putting forward. And I think the fact that three part 7000 series came out from HP has really been a major contributor to the collapse of the high end market into the mid-range. And that change is I think going to continue to go on for some time. And it benefits HP because out of all of the major vendors in the last three reported quarters we are the only major vendor to grow our market share in the external disc market. And particularly in the combination of the mid-range and high end segments. And that's because we started this disruption of the high end market and it plays right into the strength of HP's three part product. And Calvin Zito wrote an interesting piece of that. We were at Dell World and he was squinting through the numbers. So he put it that way and it was quite telling. Okay so, but that's interesting what you're saying because everybody always thought me included that the all flash arrays would eat into that tier one. You're pointing to a different now when HP acquired you and Dell wanted to acquire you. One of the areas of interest was that architecture could go from span, a lot of price ranges and that's what's happened, driven that down. That's right. And we think that there's an opportunity to continue to do that over time and continue to expand the converged flash array phenomenon bring it to new customer opportunities, open up new kind of market segments. Well, as you build these personas in for walk and file and object into three part that starts to get interesting and you know NetApp obviously has had a lot of success with its single platform, but even in its limits. They've taken on the EMC strategy, if you think about it. They have seven mode, they have cluster mode, they have ingenio, they've got a new Mars operating system. They're becoming as much of a fragmented complexity kind of storage architecture as EMC has. And they used to be the single architecture guys. It's all about time expansion, right? And you have to have sympathy for your brother. We took a different strategy though. Our time expansion is the introduction of file personas on the three part architecture, single architecture, but accessing a new $4.5 billion mid-range NAS tag. Well, this is kind of what I want to get to, is that NetApp kind of had it right until they ran out of market, right? And then hit the ceiling. Do you think that three part could become that, you know, to use a term over this term, that unified array that can actually expand price bands? What does that do to other parts of your portfolio? You know, at the end of the day, as long as we have a single architecture that can meet more and more customers' needs and deliver the business outcomes that they're looking for, if that allows us to continue to simplify our product offering, it's better for our customers, easier training, manageability, interoperability, replication for them. It's easier for our channel partners because they have less storage architectures to have to learn. And it's easier for HP because we can focus all of our R&D and innovation and investment in a single architecture rather than lots of different architectures. And that's the reason why we're so focused on in each separate domain of storage, having a single architecture that is our primary kind of go-forward architecture, for instance, for the new style of IT. So for a systems-based primary architecture, it's three-part store-serve. The software-defined, it's store-virtual. For data protection, it's store-once. And that's really what our strategy is. Yeah, well, the store-virtual piece of interest is a lot of money flowing into that whole space. We call it service-and. We've quantified that. You guys had a big lead there. A lot of people coming into that space now. Where do you think that can go? You know, I think once again, we are incredibly well-positioned in software-defined because there's a tight linkage between software-defined storage and hyper-converged appliances. What we are interested in doing is allowing people who want to build out an industry-standard server-based architecture where they can run compute and applications side by side with their storage services to be able to do that and scale either in a fixed manner with fixed compute and storage scaling building blocks on a flexible manner where they can grow either compute separately or storage separately. Because one of the big problems about conversion infrastructure solutions and particularly hyper-conversion appliances is not all workloads scale exactly in the building blocks of compute and storage ratios that come. Normally, what happens is you end up with too much compute or too much storage and the efficiency of the data center breaks down. So having a single architecture that can span from a hyper-converged appliance like our CS200-HC Store Virtual through to a VSA-based approach which can run on anybody's hardware, you know, Lenovo, Cisco, Dell, HP, anybody's hypervisor, Microsoft Hyper-V, VMware, KVM, or scale out with our Store Virtual 4,000 appliances or be integrated into Helion OpenStack as our Store Virtual software is. But have the same data services. You can build a complete data center architecture, consistent, coherent, elegant, very effectively managed under environments like OpenStack and OneView. And that is something that is not available. If you buy new tactics, you've just had a new silo. It's a hyper-converged appliance silo. You have no flexible scaling mechanism to do it. And that, I think, is the big advantage we will offer something to try to store. We've got to get going. I want to ask one final question that we've got into the keynotes. You know, Dave mentioned, you mentioned four years of escape velocity. On the startup, we watch the startup community as kind of a barometer for the industry. What's your take on that? Being you've been an entrepreneur, you had to make payroll, run down, put your credit card down. You've been there, done that. We're seeing different versions of a dot-com bubble kind of popping going on. What's different though, these are people actually making money. So you're seeing a different climate. I'm not saying it would be a bubble burst, I don't want to put words in your mouth, but unlike the dot-com bubble, there are actually violent companies out there. But is there enough room in the market for all these startups and all these valuations? Do you see that kind of getting a little bit more frothy or less frothy? What do you see in the startup world? I think we're quite honestly on some of the extreme edge of valuations and the amount of money that's been raised in the private world. To kind of raise in the storage environment over half a billion dollars and have kind of three billion dollar private valuations for a storage infrastructure company, that's slightly frothy to me. And I think the trouble will be if companies who have that kind of private valuations want to go into the public market where people are starting to look at them and evaluate their valuations based on kind of normal public market metrics, it's going to be a very, very difficult transition. I'll say it, it'll be a bloodbath. I think it will be a difficult transition. And quite frankly, there simply aren't enough acquirers left for all of these private companies to find private, you know, public company homes. So that's the big problem. When the music stops, if you're not sitting in a chair, then it's going to be difficult for a startup. That's exactly. Unless they have a revenue venture. So what's your advice to those startups? Start looking for a buyer now or generate revenue? We're hiring. Okay, we got a break from the keynote. It's David Scott here inside the queue. Always great to chat with you. Again, great, great support here in the queue. More importantly, impressive performance on HB stores, congratulations. We'll be right back after this short break. This is theCUBE live in Barcelona. We'll be right back.