 Okay, we're back live here inside theCUBE. This is our exclusive coverage of VMworld 2013. This is SiliconANGLE and wikibon.org. It's theCUBE, our flagship program. We go out to the events, extract the signal from the noise, open source, all free content. Go check out our wikibon free research. Go to SiliconANGLE, SiliconANGLE TV for all the videos. I'm here with two awesome guests, Peter Levine, former entrepreneur, still got the entrepreneurial bug in when I talk to him, and general partner, Andreessen Horowitz, mega VC firm, doing some great things, and JR Rivers who's the co-founder and CEO of Qoom's network's hot startup funded by Peter. Guys, welcome to theCUBE. Thank you, welcome. So theCUBE is a friendly place where we let ideas get out there, throw out some trial balloons, and also speculate on kind of what's happening. And one of the themes is we extract the signal from the noise. So when you have these vendor shows, less VMworld than some other ones, there's a lot of hype and positioning. Oh yeah, we're going to be the next software to find data center, all this stuff's going to be great. But as a startup, you're in the trenches and you've got to convince him for the big money. So I want to get the in the trenches view, JR, we'll start with you. What's your take of VMworld, right? And what is the software-defined mission really all about? Why is all that hype being generated around software-defined data center? Right. You know, we're a startup, you know, like you said, kind of digging through, trying to talk to a lot of customers. But what's really lucky for us is they actually talk to us and we get to listen to them. You know, there's not a lot of sales people and the way there's a lot of process and history. So it's a reasonably fresh relationship. And we're lucky enough, actually to some degree through Andrews and Horowitz, to talk to a pretty broad range of customers like a Coke, a FedEx, you know, a Microsoft, you know, up and down the stack, you know, small enterprises, mid-market, everything. And consistently, what I see in the companies I speak to is their fundamental IT problem is workflow orchestration. They inevitably have some set of apps, some are legacy, some are, you know, new rollouts, whatever, and they need to be able to deploy them over their infrastructure in a reasonably efficient way. They need the infrastructure itself to be capital efficient, whether they build it themselves or they use it from, you know, from Amazon or a hybrid cloud from VMware or whatever. They also need the cost of having somebody push an application out to be really low. And in general, it's not. It's horribly expensive because people have been growing in traditional ways for a very, very long time. In the 15 years, we've been doing a lot of manual configuration of networking, manual configuration of servers, manual configuration of storage. So the big push right here is solving that customer problem around workflow orchestration. And automation, right? Yeah, well, automation is how you solve the problem around workflow orchestration. And so if you stand back and look at it, many of the, actually all of the cloud providers have done a great job of taking care of that in their own context at Google and Amazon, Microsoft, Facebook. They spent smart engineer cycles, developed homegrown tools, and now they're in a great spot. They have very capital efficient infrastructure and they have highly automated deployment mechanisms. And a lot of the enterprises are trying to get to the same spot. So that's the software defined data center. Peter, I got to go to you on this one because Dave and Dave Volante, my co-host and I talk about this all the time because we talk to HP and all the other vendors. You know, converged infrastructure has been around for a while, that notion. But with software, with network virtualization kind of highlights that it's happening but differently than what people expected. What did you see in these guys when they walked on the door? And you have a lot of experience with Veritas and you've been successful entrepreneur yourself. What clicked for you with these guys when you walked in? What are they disrupting and why did you invest in them? We fundamentally believe that Cumulus is really changing the networking landscape. And if I look overall broadly speaking, there's a renaissance in enterprise computing and that's occurring at the infrastructure level with SaaS and mobile kind of all coming together. And at the infrastructure level, there's tremendous innovation in both networking, storage and compute. And we saw JR and team having this great background, this great networking kind of history to say we're really going to change the way networking is being done from a software standpoint in future networks as folks look to scale out and go change up their data center architectures. So the technical chops and they had to passion and they had to passion to do something radically big. Absolutely. The JR, so let's translate on your insanity. How did you convince him to give him all that money? Actually, it was incredibly easy. We had a going in thesis in that when you look at physical networks, if a customer spends a dollar on physical networking, whether it's capital dollar or an operational dollar, they want the best network they can get for that dollar, which is going to be at first, the first digit of that is all about capacity. And right now when you look at enterprise networks or even a lot of the cloud networks, their capacity bound, the CPUs are way higher capacity than the networks will support right now. And so we went in and we started talking to the partner set and Mark, two slides in, looked at this and he completely locked in on it and he stood up and effectively went through our slide deck for us ahead of time. I basically hit go on the slides because he was telling everybody how it went. And the fundamental thesis that comes out of it is if you roll back a few years, Sun kind of ruled the landscape on compute. Everybody had Sun in the data center and applications were reasonably limited. Linux came along, the commodity server infrastructure came along and businesses such as Google and AWS are the outcome of that. Nobody would have thought that was going to happen, but it did because this new thing came along. And as physical networking expands and people get more capacity, people will write applications that you can't conceive of today. You know, Peter, I met you at John Barrett's 50th birthday party, we were chatting and something dawned upon me and then you just mentioned it there. I wanted to bring it to you now is that there's a new guard coming in. It's not suddenly a new blood, young blood. It's experienced guys who think differently. But I want to get your perspective. You were very successful at Veritas. You were an early employee at Veritas. You were very successful in sense. I want to ask you two questions. What common denominator in your success was involved in those? And two, as the new entrepreneurs come in, whether they're experienced guys with tech shops or the young blood coming in, maybe under 30, what is the winning formula for them? So what's the common denominator of your experiences? Because Facebook and these new apps are defining paradigms, which have underpinnings that have enabled technical, have technical engines that might not look like yesterday's engine. You know what I'm saying? Or new tech. What common denominator was to your success at Veritas and Zensource and Citrix? And then what does the new blood need to do to have that winning formula? You know, for entrepreneurs, and this is broadly speaking, not necessarily enterprise or consumer, we look for entrepreneurs that have a secret. You have to have a secret. You have to know either something from your past or some secret that makes your idea truly unique. And so the first thing we're looking for is what is that secret? I can tell you from JR's standpoint the secret was all about unlocking Linux and making that available now for networking as opposed to in the past we saw the Sun Microsystem example taking this giant box that had everything together, got disaggregated into Linux and compute servers, well we're seeing the same thing and so the secret was that unlocking and using Linux for the basis of networking. So it's really about the secret and I can say from my past, I've gotten involved in companies where there was a secret and we had the opportunity to kind of prosecute on that. That's kind of a good BS detector too when you see an entrepreneur walk in the door, it's like if you can't see the secret, they're just copying somebody else, right? Exactly, and we look for outliers where it truly is a secret. If everyone knows about it, it's not a secret anymore and that is herd mentality. And so there's no other winning formula other than if you got to have some sort of domain expertise and the ability to pull it off. Obviously, it's a technical. Yeah, execution is huge and then market timing is very important. The right ingredients need to come together but fundamentally it's about a secret. Such AR, obviously the Linux revolution has been great. We've seen that going on, but no one wants, the new guys and the stuff that we're doing with our cloud development we're working on, no one wants to download patches anymore. No one wants to upload. They want open source, they want free, they want freedom. They want free software, don't mind contributing on a community, but they want freedom. So when you get into this network virtualization, it kind of changes this computer science paradigm but also networking, a lot of the configuration management provisioning, these are hairy technical issues. So what's the big secret on your end outside of unlocking Linux going a little bit further? That's a good question. I'm trying to pull that together. You said a bunch of different pieces and let me just give me a second to let them register and pull them together. You talked about downloading patches, using configuration management tools. I think that the fundamental secret, it's actually kind of obvious where we see it. Right now in a modern data, not a modern data center, a traditional enterprise data center, there's two centers of operational expertise at least, one of around compute and servers, another around networking. And when you look at that, the compute people have had a large number of nodes and in general, they're extremely operationally efficient. And in general, that's not true of the networking side of the house. And so inevitably the big secret is, how do I leverage the tools and knowledge I have on the compute side of the house for my networking equipment? So the secret there that we came to was, you should be using Linux because it's the same thing people are using on the compute side of the house. So again, not necessarily rocket science, but it's making that big jump that, hey, this really solves a massive customer problem that they need to address as they make these transitions. Awesome, Peter. We were talking last night briefly. We saw each other at the W about compute. You have a vision on compute. Can you share that vision? I mean, this kind of came up a little bit with the Mark Andreessen, Pat Gelsinger conversation with Andy Besenstein with Paul Sappho yesterday, the future of IT. I think you mentioned compute is going to be commoditized or what was, how did you phrase it? Give your vision, share your vision. I think the world is moving to, and if we look at Google or Facebook as examples, the world is moving to an environment where there are lots and lots of small servers aggregated together to provide a compute, storage and networking infrastructure for users. And so I think that fundamentally changes how the future of the data center and what that looks like. What's interesting to me is that Google and Facebook and Amazon are defining the blueprints for future data center architectures. And what we're seeing is that on-premise enterprises are starting to copy those blueprints for their own implementation. And so I see these environments having thousands, if not hundreds of thousands of aggregated servers all attached to all containing storage, having a networking infrastructure that's very different than the environments that we're coming out of right now. Yeah, and you know, the DevOps early pioneers were like eating glass, doing all everything, you know, vertically integrating, and there was no real operations and all engineering, and you know, Facebook's obviously the poster child for that. But DevOps is now a cultural thing in enterprises because you really can't take a DevOps guy and most of the guys in IT in Iowa, some end company, I'm a DevOps guy. No, no, you're an IT guy. I want to be like DevOps guys, but I need DevOps functionality. So that's a trend that we're seeing is the key of the virtualization piece. And I want to ask you guys both, if you can answer, is we're trying to figure this out, is where's the bottleneck in that? So is it the network? Is it the compute? If an IT shop in Iowa wants to go to the new model, copy Google, they're going to have to dial up an order and say I want some Google infrastructure? No, I mean that's what the trends are all about. You know, you asked about software defined data center. VMware is trying to lead their way with their cloud services. You have OpenStack on the other side. The difference is someone like a Google will go build this for themselves. Other people, like your person in Iowa, your company in Iowa, wants to leverage commercial offerings, they get them to the same spot. And so they've been leveraging commercial offerings for a really long time and putting together with manual glue and barbed wire. And now they recognize they need better efficiencies at scale. Awesome, and the big trend obviously, Pat's pitching the data center and I said hey, software defines great, what happens to the data center? So the question I have for you guys, just kind of an industry perspective, the data center's not going where the physical asset is still going to be fundamentally re-changed, but still we need more power and cooling, we need more drive, store stuff. So big data, drive and all that. So software defined as great with virtualization, but what's the data center look like? What's your perspective and around the trends around the data center? I'll see space issues, power and cooling. Power and cooling are huge issues when it comes to data center. And that's why we really believe that this whole scale out architecture helps to solve some of the power and cooling issues. We have this belief where the building becomes the server. So we move to this architecture where there may be a million servers inside a building, the building becomes the server and that's the new footprint and architecture for the data center. And it's very different than what it looks like now. So again, we see a lot of enterprises taking the architectural blueprints of an Amazon and saying, we're going to go build that internally. You know, I get excited when you say that because the tech trends around that concept is mind blowing because I went to school about an operating systems degree and I was just having a conversation with an analyst and he studies systems management. I go, how do you manage systems? How do you monitor systems management when the operating systems are changing? And the systems aren't software, it's data centers, it's buildings, so they just to grok that concept, how do you manage that? Is it the printer as a server? Is the phone, is the device on the person's body, the Google Glass as a server? So when you think about thousands of servers, I mean, it's just interesting to just kind of think about what's the path of North, South, East, West on the network side? So I mean, it's kind of mind numbing there, but I mean, in that world, I just don't see anything out there that can do that. I mean, what are you seeing for solutions that could create a thousand servers? Is it network virtualization? I mean, what's your vision on that? I mean, network virtualization is a tool to the end, right? Inevitably what it takes is putting together lots of small, well understood elements in a way that you can leverage on top of, right? So if you take servers and deploy them and do the physical level monitoring with a set of Linux-based tools, you can deal with all the environmentals, power monitoring, application health and that sort of thing. Then you, on top of that, you can layer something like network virtualization to distribute workloads across that substrate of compute in a way that's a provisioning layer. So it's got high-scale and easy mechanisms for monitoring. You can use OpenStack or vCloud Director as a way of do provisioning and deploying the actual workload and it's storage back into resources. And you connect them in the way that Peter was saying. It's really about a network substrate that connects together storage elements and compute elements in a way that people can provision. That's awesome. So let's get a plug-in for your company. Give us a status on stats, a funding if you have disclosed it, employees, kind of the customer base. Can you just share some stats with folks and who you're looking for to hire and what's the plan? Oh yeah, we're hiring great engineers if you're networking or Linux, sign up. Customer experience people, we have some really great DevOps people on staff that help our customers be really successful. As far as other stats, we just launched a couple months ago. We've got almost 20 customers now that are actually deployed into production. We have a bunch of interesting group of concepts going at some pretty big enterprises, all good stuff. Andreessen was our lead investor, battery ventures was behind. We've raised about 50 million in funding and we're going strong. Peter, let's talk about Andreessen. Obviously, enterprise is all the rage now, all the VC circles, I'm living Palo Alto, so you bump into everyone, oh yeah, we're in the VC sector, it's hot in the enterprise. I mean, come on, it's like most of Punchline. You can't just be a good VC in the enterprise. What are you guys up to and what's the secret in the enterprise? Well, we look to entrepreneurs for the secret, right? It's not us to come up with the secret. We look for the entrepreneur to come in with the secret. So if anybody has some great secrets, we're open for the consumer enterprise. My secret is picking the right secret and we have one right here. You guys doing great, big fan, Andreessen Horowitz, you guys doing a great job, changing the rules. I mean, talk about DeFi convention. You guys got in-house personnel, added value, founder friendly, great firm. We really appreciate that disruption. So this is theCUBE, we'll be right back with our next guest, instructing the signal from the noise here from the entrepreneurs, here from the VCs. A lot of action here at VM. We'll be right back after this short break.