 Okay, welcome back to theCUBE. We are live here at VMworld 2013, here in San Francisco live at the Moscone South Lobby. I'm John Furrier with SiliconANGLE. John Wood, Stu Miniman, analyst at Wikibon. And our next guest is Pujon Kumar, CEO, co-founder of Pernex Data. Welcome to theCUBE. Thanks John, thanks for having me here. So we love talking to startups. We love talking to entrepreneurs. It's always the best interviews because there's never a question we cannot ask. Like, hey, what's the funding like? What's it like being a startup? Everyone loves startups because it's hard. Even Paul Graham, today of Y Combinator, wrote a post that said he would never do a startup. It's just too hard. Most people don't do startups. But most guys who are crazy enough and smart enough to do that, thanks for coming on theCUBE. So the first thing I got to ask you is, what's it like here at VMworld for you and your company? And in the ecosystem, all the buzz around cloud style, hypervisor award, we just had an open stack conversation with Rick Jackson from Rackspace. I mean, talk about what it's like for you and your company here at VMworld. This is, again, for our company especially, this is an exciting time for us because this happened only once. We, this is the first VMworld, obviously. We came out of stealth just about in a few months back and we are squarely in the VMworld ecosystem. Like, once we go to the cloud, you'll know more. But essentially, everything that we do is centered around VMworld. So this is the biggest event of the year for us and happened only once for us, the first time. For the folks out there, talk about the company, quick fact-toids on the company's size, funding, customers, product, and then we'll dig into some of the tech. Fantastic, sure. So we've been around about a year and a half now. We started in Feb 2012. Two of us, myself and my co-founder Satyam Vagani, both of us came from VMware. My background has been built Exadata for Oracle, so entrepreneur by heart, and built a very large business for Oracle. And then went on to do a startup before Pernix, spent two years at VMware, and then here you go, right? And in the last one and a half year, I think we have been the fastest growing enterprise storage startup out there. Building software, we have 55 people now. We raised a series of financing when we started from Lightspeed Ventures, and then we went on to do a series B just recently, three months back from Klein and Perkins. So backed by Lightspeed Ventures and Klein and Perkins, 27 million raised to date, and also backed by some heavy-eaters like Mark Leslie, John Thompson, Lane Bess, folks have really built very large companies in the past. And they know the enterprise space. So Stu, I want to get your take on that, because we've been following these guys. Stu, John, one of the things why we wanted to bring you on is you've got great background. As you said, you worked on Oracle Exadata. Inside VMware, working on their data services and now in a startup. Can you talk a little bit about how we've seen architectures changing and how Flash especially is changing architectures from Oracle through VMware into today? What led you to Pernix? Yes, so again, it's like this, right? I built Exadata for Oracle, right? And it sold squarely on one thing. It was about performance. The traditional sands were not able to keep up with the typical data warehousing workloads. And we built a custom-built appliance. It had gobs and gobs of flash inside. Nobody knows that, but basically a purpose-built appliance that gave 10 to 30X performance improvements for that particular workload. It's called Big Data now, back in the day data warehousing. And now you move from that physical world of Oracle databases to a virtualization world where we all know that virtualization has taken over the data center. And what has come in the way now for people to virtualize more and more? It's really coming down to storage, right? Can the sands of the world keep up and how do you really solve the storage performance that has been created by virtualization? And that's one of the things that led to the birth, I would say, of Pernix data. Now having done this before, in the context of Oracle where I took compute and got it moved to storage, it was time to do the reverse with virtualization and Flash and get storage close to compute. Yeah, so if we look at Oracle, the thing that they've done really well is they built the RedStack all the way up through the application and it's optimized, hyper-optimized for that technology. Virtualization in many cases is used for kind of a general purpose environment. Wikibon CTO David Floyd just published a piece talking about the value of Converged Infrastructure, said the higher you go up the stack, it's actually kind of exponential, how much value it adds. Can you talk a little bit about your thoughts as to how architectures fit and how important the application is? Yeah, so basically it's like this, right? So fundamentally what I believe is the virtualization is at the end of the day, it's horizontal play, right? And it's really about coming up with a solution that can fix from an infrastructure perspective, do things. And these are some of the reasons that we did Pernix data for, which is at some level you've got to build an architecture that is agnostic to the application running. But there's always the one person of the use case that will not fit in, but how do you build something that is not purpose built? Like, you know, we built for Exadata and then Oracle went on to do Excelogix and God knows what. But bottom line is how do you build something that can be agnostic to the application that's running on top of me? And how do you build something that can scale out and give what today's applications need from a performance perspective? Great, can you explain Pernix data? What's the scale out story? How does that change the discussion of flash from what we've been having today? We had some server based flash, we have some violin just announced an IPO, kind of the appliance base, there's all the arrays coming out. How does Pernix data kind of change the flash discussion and where does it fit in the ecosystem? Yeah, that's a great question. So essentially, look at what we're seeing out there. So fundamentally everybody is trying to say, how do I really solve the storage performance problem that has been created by virtualization? And when we looked at the market, we saw why do you need to solve storage performance in a way that requires you to ship capacity? Why do you really need to solve a storage performance problem that requires customers to replace their existing infrastructure? Why can't we do something in software, right? And everybody talks about software all the time, but at the end of the day, they ship it in the form of a custom-built appliance. We talked about X-rated software too, but the reality of it was, it was software, sure, obviously it was software, but there was a reason it was shipped as a appliance because there was no other way around it. But why do you need to do that? How about we give, think of this, rethink this whole thing in the context of Flash, Flash in the server, and say you keep your existing storage infrastructure for what it is good at, and it is very good at capacity and data services. Keep it for what it is good at, and solve performance in a scale-out manner on the server side. And that's what we went out to do. We said we'll solve performance with a software solution, a disruptive technology, but totally non-disruptive for adoption from a customer perspective. Similar to what VMware was back in the day. If you look at VMware, it did not require applications to change. You could run your Microsoft application, whatever it is, as is. So non-disruptive, but a totally disruptive technology. So that's what we built. We built a software platform that clusters Flash in the server across a bunch of servers, and basically builds this tier that works as a data acceleration tier on top of existing storage infrastructure, existing servers, and existing applications. And that's what we did. And so it's a piece of software that goes in the hypervisor and does a scale- So it's not a hardware solution, it's all software. It's all software, and we wanted to impact every data center out there. And the only way to do that is to build a true software solution that supports any server, any storage, and any application. We interviewed D-Raj yesterday from Nutanus. He was also funded by Lightspeed, a little Lightspeed mafia in the storage business going on these days. I don't know the right talent. But one of the things he said was, you know, early on there's a lot of naysayers, because he was doing something different. You guys have a similar approach. I want to ask you kind of a personal question, then a business question. When you were getting funding, obviously you have a track rig with Exadata, you probably got a check just on your previous experience, but you had to share the vision. What was the vision that you shared then, and what's changed now, if anything, in terms of your execution for the marketplace? So I guess I would say that we have been fortunate. We, obviously, from a check perspective, you're right. Based on the backgrounds, we're like, okay, these guys are saying something. You know, these guys can really do it. But for us, what happened is, we came up with a vision. We said, you know, we got to give scale out storage to everyone out there, you know? And the reality of it was, there's one thing to do it in PowerPoint. I'm sure there's a whole bunch of folks who'll be saying this, but there's something different to give it in a product that works like a charm. So think of, when you look at, have you ever created a slide show, for example? You throw, you know, photos in a slide show application and it creates this beautiful video for you, right? So that's the kind of experience. You throw, you know, flash, you throw your virtual machines, and it basically gives you exactly what you need. So that's a vision we had, but making a vision. Self-figuring, self-assembly infrastructure. Exactly. That's what you were saying. And so taking that vision into a product, an enterprise product, that can be applicable to every data center out there, was what we were focused on doing for the last one. So did you have some naysayers? We were actually, it's interesting. More sort of naysayers. Some of the folks who questioned us in the beginning have turned around and said, oh, this really makes sense. They got it. As soon as they saw the product in action, they said, wow, for the first time I can, I don't need to think of capacity and performance in a monolithic tier. I can decouple these two things. And there you go. We have our- They all clicked from there. They drank the Kool-Aid at that point. Pretty much. As they say in Silicon Valley. Still, you have a question? Yeah, so Pujon, you know, VMware is, you know, not being quiet when it comes to storage, when it comes to Flash. Last year, they previewed what they called V Flash. And this year, they unveiled it. I believe it's like, you know, Read Flash, Cache, something like that. But in the hypervisor, it's a feature that goes there. Does this really, how does that compare to what you're doing? Does it replace it? Does it complement it? You know, how does that line up competitively? Great question. So again, it's like this. The whole market that deals with niche solutions, read-only, single-host has been around for some time, right? You know, Infusion IO has had something there. And there's a whole bunch of companies who are doing something there. But the reality of it is what customers today need is something that they can use it as fundamental infrastructure. In fact, VMware obviously is a great company. We're all here because of that. But they have violated one fundamental thing by doing the single-host solution because VMware is fundamentally a clustered hypervisor. Enterprise customers are used to using it as a clustered hypervisor. And that's the thing that's missing. You've got to build something that's clustered, that's not just read-only. And Flash in the server is in its early stages, right? People do not know, if you tell somebody that I want to statically partition my Flash and give it to each VM, they don't know how much to give it. They actually don't even know how much Flash to put in sometimes. They ask us all the time, how much Flash should we put in in our server to really benefit from it? And VMware solution, for whatever it's worth, is a single-host solution. It's a niche read-only use case. And it requires to statically partition Flash. And there's a bunch of other limitations. So fundamentally, what I see is, it is a big limitation from an adoption perspective for the enterprise. Whereas something like Purnix is what enterprise customers are really looking for. Yeah, so absolutely, one of the biggest limitations on Flash in the server was that single server. You're saying VMware yet doesn't yet do that. You have it. If I look at the Flash, and even larger kind of converged solutions are growing, huge growth rate. How does your solution fit into VSAN or VBlock or any of the other converged infrastructure solutions out there? So there are some places where we are relevant, some places where we are not. Where it's a closed architecture, obviously we're not relevant today. But in things like VBlock, totally relevant, right? VBlock is looking to put Flash on their server side. And when you put Flash on the server side, how do you really do what we are doing? They need a piece of software like ours to cluster that Flash in the server side. So we are relevant in some places. We are not. Poojahn, thanks for coming on theCUBE. Always great to have you. I wish we had more time, and we will definitely follow up with you. Where's your office in Silicon Valley? We are in San Jose, the best in San Jose. Okay, perfect. So we're, the Palo Alto team will get with you. Worth doing a drill down. Very disruptive. Again, like Newton, some people just don't get it until they see it. So we really appreciate your time coming inside theCUBE and we'll do a drill down on theCUBE. But then Palo Alto will follow up. Thanks for coming on theCUBE and sharing your experiences. Congratulations on the funding. Thank you. And we'll be following you. This is theCUBE. We'll be right back with our next guest. We're after the short break. All right. I had to kind of reset my mind there. Thank you. Going to Disneyland. Going to Disneyland. I mean, these guys are great. I think this is a revolutionary forum. Up till a few years ago, I'd never seen this in my...