 Live from San Francisco, California, it's theCUBE at VMworld 2014. Brought to you by VMware. Cisco, EMC, HP, and Nutanix. Welcome back to San Francisco, everybody. This is Dave Vellante with Wikibon.org, and this is theCUBE SiliconANGLE's continuous coverage of VMworld 2014. This is our fifth year covering VMworld. We got it all started in 2010 at Moscone in this actual very location. And so we have always covered sort of the disruptions in the market and have focused a lot lately on storage and what's happening with Flash. In fact, in 2010, we were talking about it. It had not nearly taken hold as much as it has today. Ehud Roka is here. He's one of the founders of ExtremeIO, a CUBE alum. Ehud, welcome back to theCUBE. It's good to see you again. So this is quite an event here. We last saw each other, I think, at EMC World. Right. Little different vibe here, right? A lot more of your competitors, cooperators, and so forth. So you get a wider industry perspective. You guys, we're talking off camera, let's talk about how you guys are making an impact in the marketplace. Yes. We talked at EMC World about, a lot about the journey that you guys were on, how you got started, and then you come into EMC and you get EMC eyes, and you get this huge channel to take advantage of. So it's starting to work. David Goulden has shared some numbers. It looks impressive. I've said publicly, no matter what, because everybody's claiming number one. So no matter what, EMC in my prediction is going to be number one in this marketplace because you have the channel and you've got the relationships. So how is it that you're taking the market? What kind of progress are you guys making? Give us the update. Yes, absolutely. So yeah, we talked just a few months ago and we were talking about integration with EMC and kind of an early business creation. And just a year ago, VMware of last year, we were during our directed availability phase before general availability. So within this month of less than a year perspective, I think we went from market penetration, very fast growth into what's becoming a mainstream business that demonstrates unprecedented growth. I believe we are showing growth rates that are unheard of in this infrastructure market. Okay, so I got to ask, I have to follow up with that. Now obviously you can't share the details of the numbers, but that's an interesting statement that you just made. So are you comparing it to spinning disk, other storage products, other infrastructure products? I think- History of IT? Yeah, absolutely. So I'm not a historian myself, but comparing it to other storage products, other IT infrastructure products, different waves of new technologies being introduced. I think the old flesh introduction that's happening last year, this year in a very big way, is bigger, and it's quickly transitioned or proved itself to be much more than a small, nichey segment at the tip of this performance triangle. It's much more of a mainstream, general purpose, all flesh storage that with the advancements in data reduction, economics, efficiencies, proves itself to be delivering the right economics to become mainstream. Well, at EMC you've got a couple of reference points for successful products. I would say two notable ones. It would be the original Symmetrix, which I have a long story that I won't tell you about here, but I saw the ascendancy of that product, which was unprecedented because at the time IBM had a monopoly. And then of course the DG Clarion at the time which EMC bought and then really started to take hold in the mainstream, others would be NetApp. I mean, you could track that growth. And so you've looked at those and you're saying that the extreme IO adoption is higher than anything in modern history. I believe that to be the case. I think it exceeds expectations. We won't hold you to it, but it's big. Don't hold me to it, it's big and fast. And we had very high expectations for ourselves and we are exceeding our own expectations, yes. Why do you think that is? I think that is because there is a real need that we are addressing on agility, on performance, on simplicity, on economics, on scale and the product works extremely well. And the combination of a product that was introduced only recently but is proving to be extremely mature and with the demand we are seeing in the field, these two create the success. So I have to ask you about, when I go back to EMC world, I like the conversation we had. Obviously the folks that I talked to from extreme IO are extremely focused on that all flash product. But the high level messaging was, I thought a little bit of a hybrid to steal a theme from this show. Kind of, well, there's going to be still a lot of spinning disk around. IDC numbers show by 2017, a small portion. I talked to a lot of customers who said, seems like EMC kind of trying to hold on to the past. And then very quickly I've seen a change in the marketplace. You guys have become much more aggressive. You guys, not extreme IO side of the house, but EMC as a company, much more aggressive about leading with flash. Is that an accurate perception and why the change? I think it is. And it's not to say that disk hybrid solutions have their place maybe forever, maybe for many years to come. But I think EMC has proved that it can be disruptive. It can lead in a disruptive market. It can disrupt itself in places where it makes sense. So EMC is absolutely leading with extreme IO wherever it makes sense. And it is making sense in a growing number of use cases, workloads, and across all verticals. So how should we think about extreme IO's entrance into the marketplace relative to both existing platforms, in particular VMAX, but also VFX, and early competitors? Talk about that a little bit. When I say early competitors, I mean all flash competitors. Yeah, absolutely. So again, I think we find ourselves competing in the marketplace much more with other all-flash array players than with other EMC products. Not to say there is no overlap, but overlap is good I think as EMC strategies is proving. Like Joe Tucci says, overlap is better than gaps. Better than gaps, and it's real. And I think from my perspective, fighting for extreme IO attention in the market, I think the approach where let the customers decide what's good for them makes a lot of sense. And we are fighting much more with other all-flash arrays. I can say we are winning at the vast majority of times, so our win rate statistics are amazing. And we see some overlap with other products, but it's a small portion of our... Why do you win? If I have the win pie, is it product, existing relationships, trust in EMC, other factors? So obviously it's a combination and you will find examples from all of these. But I think what makes it, and tying this back to your earlier question of why is it being so successful and growing so fast? It's because we have all of this together. We have a superior product, superior architecture, and we have the EMC, not only brand and relationship, but real infrastructure. So the EMC services organization, the EMC quality that we are leveraging to deliver quality with and newly introduced product, Extreme IO, and combining this with superior offering gives us this huge success. It's okay, so you say it's a superior product, superior offering. Some would say, okay, well, it's late to the market, maybe it doesn't have this, maybe not as much data reduction. How do you respond to that? Sure, so late to the market, it's an absolute fact that we were not the first to introduce an old flash array. Not late to the market though. I think we came right in time when the market is starting to grow in a significant way, and we are absolutely riding and leading the growth curve in the market. So in no way late to the market, even though not first. Broadly speaking, right? And in terms of features, we were not first with all features. Absolutely bridged any major missing features that we thought we had. We had a major launch in July as part of EMC's mega-launch, announced snapshots, announced compression arriving very soon. And I think with that we complete the data reduction story. We complete the set of capabilities in a basic architecture that delivers advantages relative to alternatives. So I wonder if we could talk, so once you have the data compression piece there, now it becomes much more cost-effective. Is that right for customers? That is true, in a wide range of workloads it does, so early days, and by the way I agree with you, I just finished in the first half of this year, thank God it's over, but a many, many city tour, talking to customers and it's great. It was a great opportunity to talk to customers a lot of time on the road. Probably talk to five or 600 customers and when I ask them who's really adopting Flash and I dig into it, it's very early days. Most people are just putting Flash into existing arrays. Very few are absorbing all Flash arrays at a very high rate, so I think you're right on. In the early days of Flash, very early, you would hear things like, well, you shouldn't focus on cost per bit, you should focus on cost per IO, obviously that's where the advantage is. And I almost feel like that was one of those, if you can't fix it, feature it type of things. Are we with these new data reduction technologies and lower cost EMLC, getting to the point where it's really, Flash is becoming competitive on a cost per bit basis, and are we going to start talking about more along those lines? Yeah, absolutely, and I agree with that. I didn't really like the cost per IO thing. I think it's relevant in a very small subset of the use cases. The real thing is how much will it cost me to support an application, to support an environment? And in this regard, it's a combination of your performance needs and capacity needs. And with advanced data reduction, deduplication, compression that are always in line, always active, always available with consistent performance, you are able to leverage data reduction in a way that delivers value. And I think the maturity of Extreme IO today allows it to prove itself economical at a price or cost per solution, per application level, and not limiting it to the very high IOPS use cases that allow for very low attractive dollar per IOPS number. It's both of these. I wonder if we could talk about the importance of stack for a minute. I mean, storage has always been a challenge. Companies spend the better part of a decade building out and hardening a stack. And data services and quality of service and all those things that customers have come to expect, especially at a company like EMC where you're selling into the most mission critical workloads and applications. Most people would say, well, the emerging flash array, all flash array companies, Extreme IO included division of a company, don't have that robust stack. Is that a fair criticism? How long will it take to build out? And what limits does that place on that product or that classic product? Right, I think I would separate robust and complete set of every feature, every bit that exists. Robust versus comprehensive. Versus comprehensive, so easy to robust. I can say with the highest confidence, Extreme IO is robust and you don't need to believe me or believe some architectural talk, the results in the field prove that Extreme IO is robust and delivering levels of availability and reliability that put it alongside the most reliable platforms in EMC's portfolio. Is it comprehensive? It is comprehensive in everything that's highly important. There is room to grow, there is no way we, and I would say the same on alternative platforms that are only being introduced to the market. We cannot deliver day one or year one. The full set of capabilities that have been developed for Symetrics, for example, for more than a decade. But can we deliver year one, a complete set that delivers new types of value that are unavailable with traditional platforms? Absolutely yes. So we talked earlier about the time to market. Is first mover advantage in the technology business less important perhaps than maybe many people thought it would be? I mean, in social media, I think about MySpace and Facebook who ended up dominating that. Friendster, people remember that. But why that's an enterprise tech? Yeah. Good question. I don't know that there is a rule here. But I think we have seen more examples of the very, very first mover, not necessarily being the number one or the dominant player than the other side of things. So for us, not being first, but being earlier rather than later in the market with a better thought out architecture, with a better solution, and being part of the EMC organization, the relationship, the services, the infrastructure that supports us, this combination is proving better. All right, we have to leave it there. Thank you very much for coming on theCUBE again. And congratulations on the success that you've had. Best of luck we'll be watching. Thanks very much. Thank you. All right, keep it right there. We'll be back with our next guest right after this. This is theCUBE. We're live from Moscone in San Francisco.