 Live from the San Jose Mechanery Convention Center, it's theCUBE at Open Compute Project US Summit 2015. Okay, welcome back everyone. You're watching Silicon Angles theCUBE. This is our flagship program. We go out to the events that extract the civil noise. I'm John Furrier, the co-host with Jeff Frick, my co-host GM of theCUBE. Our next guest is Cole Crawford, founder and CEO of Vapor I.O. Just made, came out of stealth mode, former executive director of the Open Compute Project and founder, welcome to theCUBE again. Great to see you. Okay, so founderitis is a term that's been kicked around. So like that means you're kind of like, you do something, you find, you get a bunch of guys together, you get itchy, you scratch an itch, you founded with the team here, Open Compute Project. Okay, it's flying off the shelves. It's in, you know, it's got cruising altitude, it's still a little turbulent, but you get to start up. What's going on? Give us the quick update. You're no longer involved. What's the title, your role, your startup, what happened today? What's going on? Yeah, so first off, mouse traps and deep parks. You know, there's always kind of the one hand to hold, one throat to choke. You know, I like to take the positive spin on it. And, you know, I like to think that every now and again, you've got to stop being a preacher and you got to start, you know, being a shepherd. And so from that perspective, you know, I found that we were in a unique position and one of the nice things about working with the nonprofit ecosystem is that people want to help the industry. I mean, people all want to see the industry move forward. So, sometimes people tell you things that, you know, offer you like 10 bits of insight. If you're paying attention about, hey, this is what my pain points are. So over the past, you know, nine months, I've been thinking about what can I do for Open Compute that would move the needle. I mean, obviously the hardware side of this, the IT infrastructure side, you look at how Amazon and Rackspace and Google and Facebook are all driving that down to zero. I mean, the actual race to zero is a very real thing. I mean, don't think for a second that Jeff Bezos isn't going to drive that infrastructure cost, that public cost down to be as competitive as possible. So you have to look elsewhere to innovate. And, you know, Isaac Asimov in 1676 said, you know, if I see farther, it's only because I'm standing on the shoulders of giants. And I think Open Compute did that obviously. And I think that we're able to do that in part because we get to see everything. We get to see the trends happening. We get to see where money's being invested. We get to see where ecosystems are gaining traction and not, and to that end, you know, we were very fortunate to vapor, very fortunate to see a real need for treating infrastructure more like cattle and less like pets. So how much funding do you guys have? Who's the investor? What's going on? Because look, one of the things we're seeing, Jeff, we've always talked about is complexity kills, but also as an opportunity. It's a wind of change, consolidation, commoditization. It's okay if you're shifting the value. So that's the wind shifting analogy, right? The wind shifting value. Commoditization's okay, because that's, you know, you're not trying to squeeze money out of a low margin. Probably so, okay, race is here, I buy that. That's not a bad thing. If you're on the other end, catching the value, wind. So what's the wind at your back? What's the itch that you're scratching? Who's funded you guys? So on the funding side, we actually haven't announced yet. We're healthy, we're in good shape. I don't know if you saw the news, but we've launched a strategic relationship with Jable Circuit. Jable Circuit and their division, SAC Velocity, has initially an exclusive license to both manufacture and deliver this. So, you know, one of the cool things about Open Compute that I don't think has really caught on with the ecosystem at large is a software company like us gets to look like a hardware company. So I have no manufacturing costs. I mean, I can say, hey Jable, we'll take care of all of those like really expensive CapEx investments. So when I'm raising money, I don't have to raise hardware company money. I can raise software. You don't have the hardware costs, but someone, your partner picks that up in the relationship. It does. So there's obviously, of course, with the industry average being five to 10 million dollars per megawatt, if you want to talk to sales in our bag, let's just talk economics. I mean, it's pure economics. It's five to 10 million dollars per megawatt to build out a data center right now. So you hear all of these C-level executives saying, if you're going to do one to two megawatts of anything, don't even think of Greenfield. Like just go get a Nicola. Well, if you want to, you know, so that economics of cloud, when we can deliver you that experience and we can deliver that at half the cost that you're paying, now all of a sudden when you want to start thinking about building a data center, that prospect isn't so scary. I mean, look at Uber. They don't own any cars. Their innovation is a service-level innovation. So that brings the question, software is that new enabler? There's a lot of efficiencies with open compute and or pre-existing partnerships. So let's talk about the trends that you're seeing. We always say the data center will be the thermostat soon or the closet size. No longer monster, modular, huge data centers. I want to, if I'm on an oil rig and I want to power all my internet of things, I'm in the middle of the ocean. I need a data center. What do I do? You gave me goosebumps because I was talking to somebody yesterday and they were just kind of like, what's the pitch? What do you guys do? And so at the end of it, he's like, oh, so you're like nest. And I'm like, well, you know, I mean, we're not like nest. We're nest, but a data center. Yeah, it's not that easy, but it's a little harder than a dial, but I think you mentioned the internet of things and you talk about hybrid clouds and where you are in public cloud, versus managed hosting versus on-premise. And I mean, the reality is we've got kind of a tale of two clouds building and forming. You've got the cloud we know and love. We've got the general purpose. We put our stuff there. We trust our secure stuff and our on-premise stuff, you know, slightly more mission less critical, but still mission critical. We can put in managed hosting. And then for all the stuff we don't necessarily care about that's expensive, we put in public cloud. There's another cloud. And the current cloud is powered by your big hyperscale guys. And look how much money Amazon spends. Look how much money Google spends to allow that. Azure is now the table now. And Azure, I mean, they're taken off, right? But there's another cloud forming. And that cloud is the IoT cloud, or more accurately, probably the internet of everything cloud. I mean, it's estimated by 2020, you're going to have 40 billion connected devices. And this IoT cloud is delivered at the edge. It's not delivered in these big hyper, you know, megawatt, 100 plus megawatt. So it could be peer to peer. It could be all kinds of different protocol algorithms, blockchain and Bitcoins, looking at some interesting things there. New things are forming, basically. New things are forming. And so the sale, the winds that are back here are all around, how do we actually build the software-defined data center? You know, you heard probably Steve Herrod was the first guy ever here, I ever heard say software-defined data center. And I love Steve Herrod. And I love that saying, but you can't have a software-defined data center if you're not talking to your data center. Yeah, you can't have a data center in there. There's a data center. There's actually a software-defined data center. There's actually a software-defined data center and a facility, and there's things that, you know, that power the IT equipment. I've never seen a non-software-defined data center. Yeah, exactly. Well, that's, yeah. You can see a lot of those. It's a software-defined nothing. Yeah, exactly. So, you know, we want to expose, and I think this is really where things get interesting for this industry. And certainly in context of IoT, this is about urban, you know, urban infrastructure. This is about embedding capabilities at the edge, because IoT is all about delivering at the edge. So when Sun Microsystems 15 years ago said, the network is the computer, they were right. They were just 15 years too soon. Yeah, and actually we had Scott McNeely on the cube, and he said, I should have just called it cloud. Perfect. All right, so I got to ask you about this, because now there's a lot of different nuances to your whole vision, especially this Internet of Things, small spaces, modular capacity. Everyone's concerned with the data center density. That's, you know, tired conversation, still irrelevant. But now you're talking about Internet of Things, but you're also talking about global experience. We had Facebook on here earlier. I want to connect everyone in the world. Well, outside the U.S., you've got foreign countries, third world countries now booting up infrastructure, you know, solar powered charging stations, the base stations to, you know, people. I mean, Internet of Things includes people. People have things too. So this is the whole world we're in. So what does that mean for you guys? Does that help you? Your market opportunity? Do you sell them more? Yeah, I mean, I think for us, you know, it's really about being able to deliver that experience in a consolidated way, in a practical way, that doesn't get in your way, if that makes sense. So we want to deliver L11 solutions, which in our world is kind of like the complete, you know, you can call it data center in a box, right? And you kind of see this with the modular and the container based solutions, but no one really wants a parking lot data center. I mean, people have built rooms for this and people have, you know, floors on the 53rd floor of New York or you're paying by a riser inch. So people want to reclaim and reuse that infrastructure. We just need to offer practical solutions to the problem. It's great that we can move stuff in a container and it's nice to ship. You put it on the 18 wheeler and it gets there. But if you can ship flat and you need to deliver to a forward deployed military unit or you need to deliver it at the edge for a life science company doing protein folding or epigenetic data processing, like, you know, it's far easier to just ship a pallet, build the thing and start processing. And so for us, I think that the reality is you're not going to put 100 plus megawatt data center in downtown New York or downtown San Francisco. So we've got to do more with less, but more with more. And by that I mean, we're going to have less, we're going to have less power and we're going to have less space, right? In one and one place, but you're going to have more distributed data centers. So we need to create an intelligence layer. So data centers that are running, maybe one to three megawatts are disaggregated. You know, you hear Facebook talk a ton about the disaggregated rack and believe in that fully. Look at wedge, look at F boss, look at six pack. Look at all the technology coming out of there. It's fantastic. But we need disaggregated data centers. And if we're going to get there, we need to start at sort of the out of band management capabilities that we're talking to and through today inside of our IT ops equipment. There you've got a couple hundred dollar piece of equipment that you talked to today and it's a very high touch, high value. This is infrastructure as pets. I mean, this is a very one to one relationship with our IT ops equipment. We need to start talking to racks and data centers as one holistic thing. Say, you know, what am I paying for square foot? What am I paying for kilowatt? What am I paying for this public cloud? And where can I exist outside of here? Is there a better, more optimum place for me to exist? Finally, we'll bring some sanity to efficiencies in context of workloads. Because today the standard metric is PUE. Everybody talks about, oh, we've got a great PUE. Okay, what does PUE actually tell you? Nothing. I mean, it tells you how the electricity got from the grid to the rack. It tells you nothing about the workload. So why don't we as a community expose an API that gives you the ability to plug in performance as a variable where performance can be a transaction, performance can be a page load or an ad served, and now in context of how energy is being delivered to the rack and how well your workload is performing, you can move beyond PUE to something like performance per watt per dollar that actually gives you something measurable. So the big trend that these guys in the ecosystem are tapping into is what? What is the core value proposition that's igniting these entrepreneurs and these developers? Founderitis. What itch are they scratching? Is it engineering on the hardware side? Is it software? Are you seeing any patterns that you could share? Like, what's the general itch that they're all scratching? Yeah, I mean, I really think it's about driving intelligence into the environment. I mean, I think at every layer we need to be driving intelligence into our infrastructure. It's really the big data push. The big data's been happening for so long, but finally, you know, we've got an environment and we've got enough consensus and I really think Open Compute had a lot to do with driving the ability to get to some form of consensus around how intelligence should be gathered and delivered. So I got to ask you about the business model question. Not necessarily this is a tried and true, but open source is working. Open source certainly has changed the world and it's working here in this ecosystem. What approach are you looking at and that others should be looking at in terms of collaboratively, collectively working together? Is there an approach that you think that works best? Super VC funded, you know, build a rocket ship. Hope you don't flame out before you hit the atmosphere or go inch along, move the ball when you can. What's balancing it? I think it's, you know, I think that's kind of on a case-by-case basis. Certainly, if you're going to start up a hardware company, you're going to need a lot of capital and you're going to probably, you know, have to go give away a large percentage of your company to go get that capital. I believe in, and I think the industry is a point where, look, let me take another great example. You probably talked to him as well. Great guy, lives here in the Valley, happens to be doing some really cool things at VMware now, but look at what Martín Casado did with Nisira. You know, a bunch of really smart guys wrote OpenVSwitch, gave OpenVSwitch to the community as open source and then drove a ton of value on the last 20%. So, you've got 80% as here, go innovate. If you want to compete against us, go do it. And then the last 20% was, we're going to give you the best experience and we're going to spend so much time engineering that you're probably not going to- They built a good differentiation into their 20%. Yeah, but I love differentiating on the last 20%. Yeah, yeah, yeah, it's great contribution. Absolutely. You did your thing and now it's compete on code. That's right. And I think software is the right place to do that. So, OpenCompute allows a software company like us to look like a hardware company. On the, you know, our model, we don't manufacture a thing. We're not, you know, despite having this capability and this really cool vapor chamber, well, we don't actually manufacture it. We leave that to Jable and, you know, we license this. I think it's about licensing and, you know, we've moved beyond the open core. I'm not a huge fan of open core. We've moved beyond that conversation, I think. For instance, the open DCRE, the data center runtime environment that we gave to the foundation today. It's the first, at least as far as I know, it's the first completely reciprocal license contribution that the foundation has seen, which means you're not going to end up with a standard where one OEM is talking a slightly different version of the protocol than another OEM. And in that world, the consumer wins and the customer wins and the community wins. And as long as we're doing the right things for those people, I think you're going to find traction. You're an awesome entrepreneur, inspirational. Obviously, the work you've done here is super impressive and great community. What inspires you? And the folks out there are watching and trying to figure out where to jump into this world. What are you currently getting excited by and what should people be inspired by? I'm inspired by this, right? I'm inspired by seeing so many companies that five years ago would not be on the same floor having conversations about how we tackle this one specific problem. So what wakes me up in the morning and gets me going is seeing a community of competitive companies working in context of co-opetition. And I like that word co-opetition because in that world, the pace of innovation speeds up. The amount of technology and innovation that the consumer gets to see is unparalleled to what we've seen in the past. And I couldn't be happier to be a part of this engine because it's game-changing and completely transformative. All right, talk about vapor IO. What's next? What are you guys up to? Obviously, you're coming out of stealth, the big announcement, you're revealing, kind of what you're going for, which is essentially Internet of Things Cloud, which is a range of software technology you guys have, working on hardware, open compute or whatever, partnerships, but that's small footprint type data centers. I mean, I don't want to say data centers in a box, but it could be in a, what used to be a network ID for Verizon or Telco, you can put a data center. Do I get that right? You got it right. It's not a box though. It's a little more hexagonal than that. But we like to work in hexagons, not circles. You know, a lot of us do. Why is it circle? Circle means. I mean, bigger than a bread box or smaller? I mean, what is it? Give me some size. How big is it? It's, so it's nine feet in diameter. And inside of that nine feet, you get a collapsed, a hyper-collapse data center available to you. It's 150 kilowatts and nine feet squared. So roughly 81 square feet. And in there, you get your fog, your Fezda, your Halon, your battery backups, your rectifiers, your PDUs, your fabric, all of your IT gear. And it's all delivered to you at an extremely competitive CAPEX cost. And the OPEX, you know, the models that we're running right now, sub 1.1 PUE. So not only do you get the benefits and the economics of actually being able to feasibly build your own cloud again and put that in a facility that you've already built. But you also get the benefits of the actual cost of data centers. It's not really the IT equipment, the actual costs as the power and cooling that you've got to do to support the IT equipment. It strikes me, the 80-20 conversation because if the community via open source is all working on the 80% in theory, the last 20% should be more valuable because it's the very tippy top and all the investment is really optimized around the top 20% in terms of the value versus the bottom 80% of the infrastructure. And it's hard to get there, right? And it's hard, that last 20% is hard. Anybody can go on a ski slope and learn to ski. Anybody can do that. You could probably get to 80%. Spend a couple of, you know, an athletic person. Get to it. It takes a unique person to get that extra 20%. So it's that last mile that really is tough. But it's also- But that's the opportunity, right? But that's the opportunity. Okay, Cole, we're getting the hook here. We'll give you the final word. Share with the folks watching that are learning about Open Compute. What's going on here at the event in the moment? What's the vibe here? What's this movement all about? I mean, this is the future. You know, it's funny. Mark Andreessen said software is going to eat the world, but hardware is going to power it. Until we get to some future state when you don't need hardware, hardware is going to power that software-driven world. And I think that that's what you're seeing here. You're seeing a lot of companies that a decade ago were not paying attention to software. But you're also seeing software companies that didn't know how to be hardware companies. Now look, you've got all of these hyper-scale guys practically being their own ODMs. Amazon is now their own silicon consumer through their relationship with Intel. Facebook, obviously, orders a ton of infrastructure. So you're seeing sort of a merging of hardware companies that can look like software companies and software companies that can look like hardware companies. And this is the future. All right, Cole, thanks for joining theCUBE. We'll be back the next guest. We are live in Silicon Valley. This is theCUBE at Open Compute Project Summit 2015. We'll be right back with more coverage for this short break.