 Live from Las Vegas, it's theCUBE. Covering Discover 2016 Las Vegas. Brought to you by Hewlett Packard Enterprise. Now, here's your host, Dave Vellante. Welcome back to HPE Discover 2016, everybody. This is theCUBE. theCUBE goes out to the events. We extract the signal from the noise. We got a crowd chat going on crowdchat.net slash HPE Discover. Essentially, we're documenting a transcript of all theCUBE interviews. Check out siliconangle.tv and siliconangle.com for all the news. Check out wikibond.com for all the research. Vishmal Shand is here. He's back. We're going to geek out a little bit on storage, Vish. We haven't had time to sit here and just one-on-one talk about what's going on in the storage business. So I'm really excited for the activity that I'm seeing, the momentum that you guys have. So first of all, congratulations on all the success. Thank you, Dave. Thank you. Great to be here on. It's been quite a journey. Let's go back a little bit. I just love having this conversation. A little armchair conversation. But when you think back to 3PAR, when you guys started the company in a really exciting time, but you had this vision of kind of utility storage, the storage service provider market, which evolved, of course, into cloud storage, but you really went after a very difficult segment to go after. Tier one block storage dominated by EMC, obviously Hitachi, so to a certain extent IBM. You were able to crack that market, achieve escape velocity in a very difficult time after the dot-com boom, and then land as a public company. Great, successful exit, but it wasn't easy. Yes. You know, because you had to expand internationally and then all of a sudden, everybody wants to buy you $2.5 billion exit boom and you're inside of HP. And the rest of the story is history, but it's a great one. So first of all, up until that point, it had to be an exciting ride for you. I mean, I don't know if you've been through anything like that before. I mean, this has been a great ride, right? And while I was not there at three par from the early start, I was there for some number of years, and it did go through several, we had several phases where it was very difficult, right? The dot-com, the economy, economy slowdowns, and many companies didn't survive through that. And so when we finally culminated with the acquisition by HP, it sort of started the next phase, right, of development and growth. And so far it's been a very, very good journey for us. Well, and then, again, HP, at the time, HP, Mark Hurd was the CEO, Dave Donatelli was sort of the head of storage and wanted to do this acquisition, and it was expensive acquisition, it was $2.5 billion, but it turned out to be a great IRR for HP. But at the time, I remember the rationale was, we can take this architecture and go small to large. Correct. At the time, they didn't say we could take this same architecture and go into flash, right? What everybody predicted, myself included, and I was wrong, was that HP was going to, at the time, HP now, of course, HPE was going to have to buy an all-flash player. And you guys said, no, the architecture is going to lend itself to flash. That's how it was designed. Maybe some luck, maybe some brilliance. And everybody said, oh, boy, bolt on. Right. The balance sheet is not going to support an acquisition. That's just sort of the have, can't fix it, feature it, marketing. But okay, but it turned out to be just the case, right? That you were able to transition and become a leader. Absolutely. Amazing, amazing story. No, I think we have to thank our customers and we have to thank the founders, the three-par. And I think, while HP, at that time, they didn't say publicly that, you know, they were thinking about flash. While they didn't publicly state it, they were thinking about flash and credit to the evaluators at HP, looking at the three-par architecture, they had that on their mind. And they saw that potential, right? And then again that- Long-term planning and yeah, okay. And so they did see that potential. Now, of course, there's always a long road between potential and then delivery and then between delivery and adoption, right? And there are many, many different things that go along the way of those pieces, but you know, credit to HP for having HP at that time, for having that force. Well, and you were there in the thick of it. You were in the trenches during that transition. Yes. What was that like? I mean, requirements dealing with the technical hurdles, you know, working with customers, figuring out the pricing models, right? What was that like? Yeah, you know, the journey, I think you were referring to the, when we decided to go all flash with people, right? The supplier, the man-supply, I mean, so many factors. There's so many factors, right? So from a product strategy perspective, I think the first thing we had to handle was, okay, let's make a decision. And at that time, there was some debates on, is the flash market going to be driven primarily by latency, IOPS performance, right? And only performance, right? And you really want to drive to that really, really razor edge of performance. And we had a little bit of a different view there, right? And some of us in the company said, no, you know, performance is very important. It's fundamental, but it's not enough by itself, right? The affordability is a key factor, and so are the data services. And so getting through all of those initial debates was important. And then once we got past the debate, right? And we brought it to Dave Donatelli, and you know, he supported us, right? Through that time, getting past those debates then required the hard work of, okay, flash is different, right? What do you do different? How do you engineer that difference? If you remember, we announced the 7,000 in 2012. We announced the first all flash product in June, 2013. So that was 18 months, right? So that was 18 months what the engineering work that went into three-power all flash. Now, there was a lot of good leverage we could have taken abroad because of the architecture, but that was good 18 months with the engineering. And you have come out of it with an extremely competitive, low latency product with the industries. I think I can say this unquestionably, the industry's most robust and mature all-flash stack. I mean, that's a no-brainer, right? There's no one even close to having the hardened, mature stack that three-power has with all those other attributes. Yeah, and there's one of our challenges now, Dave. I think, you know, that richness of the stack, while it's just massive strength, I think we got to do a better job of explaining that to more and more customers, right? Because as soon as you understand it, and you see the delta differences, right? I'll give you a couple that are really key. Ability to move data non-disruptively, peer motion. People just love that, right? Even if I had to move an aged data off or flash array to something else, being able to peer motion that, big, big one. Another one, peer persistence, active-active failover with transparent orchestration. People love that. The customers really resonate to that solution, right? So, the richness is the strength, and then making sure people understand it, that's work for us. So you've created, I mean, essentially a mainframe class set of data services that, you know, SMB or high-end SMB up-pricing, I mean, it's very powerful. Okay, so now we're here today, 2016, the big announcements this week were focused on density. Manish Goyalad's breakout session showed, I love the slide, it was on Twitter, showed all these Vmaxes and all these, you know, HDS boxes and then this tiny little three-par, you know, small configuration. What's the messaging around the density and all flash? Yeah, so maybe first important to step back, Dave, and look at, you know, the three vectors that we've always talked about, performance, affordability, and enterprise class data services. And so, if you dig into the affordability section now, or the affordability vector, to me that's where density plays, right? Because that great density allows you to save power, cooling, flow space, and you know, if you're a large service provider, or a large enterprise, a private cloud provider, that density always matters. If you have, we have service providers in Tokyo, Japan, and I asked them, I said, could you use a large SSD if we doubled capacity? And the answer was, bring it on, Manish. The more you double every time, the more we take on, we have growth. And you know, if you can help us either maintain density or reduce density, that's always welcome. So this announcement, we announced two new drives, right? 7.68 and 15.36. Last year, we announced 3.84, the year before that 1.92. So it's been 2X, 2X, and now 4X, right? And if you want to take a density metric, Dave, in 2012 we were at four terabytes per rack unit. Four terabyte usable per rack unit. We are now at 536 terabytes per rack unit, okay? And again, that may be short in the window, if we look at the last 24 months, 16X improvement in density, 40% lower dollar per gig. So this is really interesting because, you know, we've been on the all flash bandwagon for a while, Wikibon, and Floyer's, you know, he's got his cost curves in it. He shows the cost per bit of flash coming down quicker. Now, sometimes that's an anomaly, you see different, you know, sourcing issues that can affect that. But there's a debate going on in the industry, particularly driven by Moshe and I, right? His company, with Affinidad, is saying, no, you don't need all flash, you need some flash. And so it's an interesting, so how do you guys, I mean, I know how you guys see it, but I wonder if you could talk about it a little bit. You believe that flash is going to be less expensive than spinning disk, let's say by the end of the decade, right? In terms of usable, right? Because you could data reduce it. TCO, we don't even talk about it. I'm just talking about. Data reduction? Yeah, yeah, so what's your take on that? I mean, you guys are obviously betting on that, but are we over, am I overstating that? No, I think clearly there is this move towards what we call the all flash data center, right? And where more and more of the mission critical, more and more of the IT data center spend goes to flash. Now, I think there's still use cases for deploying hybrid systems. We've had a couple that have come on theCUBE and have explained, right? One is the example of data aging. Right. Another one is the example of, I heard this morning from a customer that said, if I have to, by regulation, encrypt at the file and database table level, okay, and it impacts my ability to compress indeed, but the flash level, do I still buy all flash? And I said it depends. How much performance do you need? And if you have to have 1,000 spindles to get that performance, it may still be more economical with a 7.68 terabyte SD drive. Now, but if you don't have that number of spindles, maybe there's a consideration here for spinning, right? Right. So I think at the end of the day, you want to be able to have all of these tools at your hand for your disposal. And I think locking you in into any one approach, while sometimes that's forced on you because of a technical reason, if you can afford the luxury of having the option, we've seen customers go from hybrid to all flash and to what we sometimes term converge flash, right? The cloud was here last year and they talked about 30% of data not being accessed and wanting to move that to a cold tier. So again, I think that this is going to continue. How about the competitive landscape? We'd love to talk about the horses on the track and we haven't done that much this week. So you guys are leading. EMC obviously still a very, very strong competitor. Of course, the Dell merger acquisition is going to be really interesting. You got NetApp made a move to go out and buy SolidFire and they're essentially positioning SolidFire for the service provider market only. They got Ingenio over here and Cluster Nontap for everything else. Hitachi I'm not as familiar with, really not sure what they're doing, although I've been told to get some good, they always have good products. So what's the competitive landscape look like? Give us the summary there. So from a competitive point of view, it seems to me that the major storage vendors are still struggling through disruption. What I call this the multiple plan Bs and no role plan A, right? And still going through the multiple offerings. I mean, even NetApp's most recent Ingenio, all flash, SolidFire, EMC, I think it was very notable that at EMC World this year, there was no mention of extreme IO. There was all flash VNX, all flash Unity, all flash VMAX. And I have to imagine that must be confusing for their customers, let alone for their sales and channel partners that customer would ask and say, hey, what should I deploy? What do you recommend? It must be a very bewildering set of choices, right? Yeah, and I left out IBM. I mean, IBM bought Texas memory systems a while back. Good product, high performance. But they belong to a class where I think data services continues to be a difficult challenge for them. They've used an approach with SVC. If you look at the market data, I don't know, it doesn't appear to be resonating as much. What's the market data saying? The market data is showing clearly EMC, HP, pure storage, NetApp playing in here, less so for IBM. Right, and pure seems to be cruising, I mean, growing, obviously still losing money, but seems like a lot of momentum behind there. You run into those guys a lot. We run into pure storage all the time. How do you compete with pure? What's the... The biggest thing we do when we compete with pure is, first of all, the data services. The fact that we can offer pure motion, pure persistence, replication, both sync and async, we see that a lot, right? The other thing that we see against pure, they're essentially a dual controller architecture and not even an active, active, active passive architecture. And in the world of Flash, controller performance is everything on your performance, right? And so whether it's capacity scale, performance scale, you know, where we beat them, there are many places where they just don't have that scale of performance or capacity. So you're willing to get in a rock fight with any of these guys? You know, I think again, we try and stay focused on the customer need, right? And when we talk to customers, we had plenty of them this week at the show, right? They tell us what they need and then to them, uptime resiliency is number one. I think a lot of the uncertainty around the future of HP, HPE, the split of the two companies is behind you now. Seems like the whole spin merge with CSC has been a real tailwind for you. Stock's at a 52 week high, right? The balance sheet looks a lot better. The Aruba acquisition, a lot of really good things lining up. You know, the other thing is, we've said it a number of times, you say you're focused on the customer. Customers have been rooting for HP this whole time. It's an icon of the industry, Silicon Valley basically founded in a garage at HP. So it's like HP is getting back to its roots of invent. I think the company's split into HP has been very positive, right? And there's clearly a lot more focus. Today, a blogger quoted this to me and he says he's seeing great fruits of innovation. And he was mentioning to me about something called Recovery Manager Central, right? Which we announced another version of it at this show and we have added additional enhancements. Another data service, right? This ability to do flat backup, right? And he was very intrigued about the possibilities with not just flat backup as well as RMC. He also recognized that with that RMC framework and sort of keeping track of these change blocks, your ability to move data now across, say, system defined to software defined. There's some interesting possibilities there now with that RMC framework, right? And some application specificities around SAP and Oracle. Yes. All right, fish, we got to leave it there. Thanks so much for hanging out on theCUBE. Thank you, great being here again. All right, keep it right there. Everybody will be back with our next guest. Right after this short break, this is theCUBE, we're live from HPE Discover 16 in Vegas.