 Okay, we're back here live in Silicon Valley in the heart of Silicon Valley in San Jose, Convention Center for Open Compute Summit. This is theCUBE, our flagship program. We go out to the events, extract the signal from the noise. I'm John Furrier, founder of SiliconANGLE, joined my co-host, Dave Vellante, co-founder of Wikibon.org. And we're here with Sunay Tripathi, CTO founder, Pluribus Networks. Welcome to theCUBE, Stanford grad, CEO, entrepreneur. Welcome to theCUBE. Thank you. Thank you, John, for inviting me. Dave and I were just talking about all the innovations we've had a variety of great guests on today, really alpha geeks, entrepreneurs. We haven't had any venture capitals on yet, because they're too busy trying to chase down all the deals. You know it's a hot sector when there's no VCs on theCUBE, Dave. Yeah, right. That means they're actually out venturing. They're probably digging up some money. Right now is the perfect opportunity for raising venture money, basically. Yeah, and it's the hottest area. So I'm going to ask you, Open Compute Summit, what's your take of this so far? Now, obviously you're a startup. You have landscape out there. What's your take on Open Compute? Explain to the folks out there, what's this all about here? Why is it so hot? Why is Mark Zuckerberg here? Why are all the alpha geeks here? What's going on? So John, Open Compute is really like, at some point programming became open. Anyone could do it. Anyone was able to do it. And at the same time, hardware was still kind of a black science. And what these guys are doing with Open Compute, they're opening up the hardware platform, right? They're making it accessible. You can get the designs you want, basically at the price that you want. There's no more NRE involved. There's no more complex infrastructure involved. You just get what you want, basically. So think of it this way, software was already very open and flexible and now the hardware is open and flexible and both meet up. So it's a perfect opportunity for someone like me. So basically it's kind of like the old days when Steve Jobs and Wozniak were like, hey, we want some of that computing power in the main frame room. Why don't we just build our own personal computer? Yep, it's exactly. In the same way, we want some of that. Right. Give me that. Yes. So that's what? Silicon, that's semiconductors. That's platform. Silicon, semiconductors, servers, storage. And for me personally, networking is now opening up, basically. So that was pretty much one of the last bastions that is now open. And that gives us opportunities that didn't exist before. Okay, so back in the old days, we used this big metaphor, the Mac 30th birthday party, which we were just talking about this past weekend when we were there. The PC disrupted the main frame, the mini computer. What does this disrupt? What is the open compute revolution and the open stack revolution? What do they disrupt? Who are the disruptors from networking to compute? So from the networking standpoint, basically, I think a lot of the networking industry has not changed in 20 years, basically. And they are getting disrupted at this point because their entire IP and what they were selling was locked into the hardware and the hardware is now becoming open. And we are now thinking of how you use networking to do applications and various other things. And that is the big disruption happening here that the networking is just opening up like servers and storage are open. So how do you see that affecting the roles within organizations? You know, you've got the server admin, you've got the storage admin. I guess the VMware admin is sort of this new role in the last decade. Network admin, et cetera. How do these lines shift or do they? So Dave, this is an excellent question, right? Because this is the crux of it. Earlier, you had so much resources going into just putting the hardware in place. The guy who helped power was the guy who knew hardware and could resource the hardware and the right hardware was where the sales were. That is all changing to a business need, right? So you as a company or organization is figuring out what is the application that's making money for you and what does that application need? And then you work backward from it. And this is where we come in, trying to go down the path of making the hardware and operating systems work for the application and be flexible to the applications. So your business needs are now governing the infrastructure and you can change things and fly in very quickly. And that is what is really bringing this new set of innovation on. You know, you hear, remember, we used to have the business process reorganization, business process design, et cetera, et cetera. Right. I don't hear a lot of corner office types talking about restructuring the roles in IT. But it seems like that should be a starting point. It is already happening. So what are you seeing there? So what we are seeing is a lot of our conversation with bigger customers and even mid-sized customers are starting from, well, they start from saving cost in the infrastructure, but very soon they're transition into here's what I need to do to make money. Here is what my application needs to do is for us to exist. And how can you actually get this rolling? So the guys we are typically talking to are the application guy, the server guy, the virtualization guys who are trying to make the entire infrastructure work for them. And that's where we come in and make it possible. And this is what OCP is driving. This is what the entire industry is moving toward. So roles of IT now, you know, the guys, the server guy and the network guys are being relegated a little bit lower. And it's the application guys who are actually taking control. I mean, everybody's in the enterprise who's, you know, working on strategic plans and pulling out the binoculars is trying to replicate or at least in some ways duplicate the public cloud, what Amazon's done with the cloud. And of course they're trying to bring that in-house. But when you get down and talk to the nitty gritty practitioners, there's a huge gap. And John, we talked about this all the way back in 2010 when we first brought theCUBE to VMworld 2010, we talked about the gap between kind of the service providers and traditional IT. And I wonder if you could comment. Do you feel as though that gap is closing? I mean, your part of your role is to close that gap. But is it closing or are the service providers, I mean, certainly the service providers are, you know, sort of doing their thing and innovating. Is traditional IT closing that gap? Maybe the leading edge guys are, but is the fat middle? No, so I think some of it is happening. Dave, look at it five years back. You had people like, or a few years back, people like Facebooks and Googles and Yahoo's of the world who knew how to do this. But it still took them a lot of PhDs to make it work, basically, right? And a lot of resources to make it work. That is not changing, right? So what we are selling is not to the big guys, we're selling what we are finding out is the enterprise, the middle level enterprise, slightly higher level enterprise, people who can't afford PhDs anymore, people who need applications to come up and be useful immediately. Like, you know, you decided to set up this cube here. Obviously you need network feed. Obviously you need to get up and speed scale. Let's say Zuckerberg shows up here and all of a sudden you're going to get chiller traffic on your side, basically, right? So how are you going to deal with that, right? We have auto-scaling. We have auto-scaling. We have auto-scaling, we've got 60 megs of streams. We're good. I think we can handle that. Love it, so to your point, though. The little guys like us, so you're selling into that mid-market, is that right? Mid-market and enterprises and campus network. You bring up a good point, Dave. You bring up a good point that we see in other markets, for example, big data, data scientists are hard to come by and so the trend is everyone's abstracting away the complexity to make data science more like an analyst function or anyone can do data science. So if I hear you correctly, you're saying what your thesis is and what you guys are doing is you're abstracting away the PhD requirements for networking to make it work. Yes. Does that's what you're saying? That is excellent. That's exactly it, John. Make networking simple. So applications, guys, keep coming in, software guys can come in and just do what they want. That's DevOps. That's DevOps. That's DevOps. So is that, can you talk about the DevOps pieces that they do? Was it big data? Was it all the above or you guys specifically in certain environments? So Hadoop and big data is definitely a possibility and we are definitely selling there. CDNs is another area that we are selling quite a bit. People who are trying to build application at scale in the internet. The other place, because some of our offering have flash-based, PCI-based flash on our boxes as well. So network and high-speed storage in there and that's actually showing up in a lot of places. People are very interested in that because FusionI are one of our partners. We can put them on our boxes and you can get fast network and fast storage very quickly and that is picking up quite a bit. Well, you're seeing a lot of the so-called, we heard Frank this morning talk, the sort of disparage converged infrastructure. I mean, it's an okay term in my mind because we're talking about putting things in a single unit. Right. And it seems in our experience at Wikibon, talking to practitioners, the guys who have designed in flash, along with the networking, are seeing some success versus guys trying to bolt it on. Yeah, we have that too. It's sort of a check-off item. It seems to be a little bit harder to deal with. So, Dave, you're so right on. I'll give you an example. We are not public yet. We haven't launched yet, basically. But from our beta program, we are heavy in revenue because people are just wanting this stuff. They try beta and immediately go in production, basically. So what we are finding out is that top-of-the-rack network, a fast network with flash in there is a big seller for enterprises and it's very programmable. So we bring the standard server-style programmability into these things. So you can bring OpenStack in there. You can bring Big Data, Hadoop in there. You are application-aware and you know what applications are going on. You can bring your services in there. So that is really the crux of it. It's so simple that anyone can do it and that's where we are coming from. So, Sune, help us talk through this, the whole software-defined networking meme. I got to ask it. A lot of people in the audience are probably rolling their eyes, but there's an interesting battle brewing between the VMware with NSX and what Cisco is pushing with ACI. I mean, for decades, we've seen Cisco as the dominant whale in this business. They got 60, 70% market share. It's all everybody talks about close, but what's your take on what's happening? How much is real? Is the disruption finally here? So, Dave, you said you won't put me on the spot and then you spring this on me. Did we say that? It's a cube, it's going to do it. I mean, we talk about this stuff all the time. We're trying to understand it, really. No, sir, SDN, if you distinguish the hype versus reality, right? Why it came on was to program the network very flexibly, right? That's what was the goal, that I want to, based on virtual machines' need, based on application need, I want to rejig the path, I want to rejig the network and be very flexible, right? What ended up being is somewhere in there, the architecture morphed into controllers and centralized controller without understanding the scale and that's where things went downhill, kind of. So we actually still are trying to do the same mantra of programmable networks, right? But we do it differently because we build a network operating system that sits on bare metal switches and it has the switch right under its control. So the programmability is very simple and we use C, GCC, GDB to program the switch. So people who are used to programming for 20 years don't have to learn anything new. It's the same programming environment. Plus we build enough applications ourselves that it's useful from day one. So SDN guys, people, I mean, we are in SDN camp as well, right? But we have a take-off SDN as a bare metal network OS that drives SDN. I think people need to understand the scale. ARAC is very powerful today, especially with what's going on at OSCP, if you look at it, more capacity, less power and lot more applications are running in ARAC. So scale really needs to come and reliability needs to come and that's what a lot of this industry is missing right now. And historically the network has been a constraint to that scale, yeah. Who's the user of your product? The NetAdmin or the SysAdmin? So we initially got SysAdmin, DevOps, orchestration guys coming in. But what we are also finding out that we actually put together making it easy for people, a CLI as well, for people to control. And when pressure comes from CIO to their NetOps side saying that you should go to merchant silicon, you should go to the commodity market. So we are getting some of the NetOps guys in because we are actually the path of least resistance for them. Because the whole controller model is very, very alien to me. You're the puppet and chef for the silicon. Basically they're managing configuration and management of networks. So it's kind of like a DevOps network thing. Yep, DevOps networking definitely is our strength. But traditional NetOps, like I said, people who still want to deal with some scripting. Did I get that right? I mean, because a chef and puppet they do different things, but you're down lower in the stack. So if I'm building out my own compute platform, I need you. Yes, you need me to connect all the compute together. You need me for networking and storage. And plus we are doing the operating system for networking and storage. So you are independent from the server side. So you're competing with Cumulus? Cumulus is just trying to make things open and commodity. We are trying to actually add networking to give more capabilities to DevOps. So more application focus. So your value proposition, if I understand this correctly, it's kind of stitched together, because it's kind of hard, right? Your value proposition is you're targeting people who are building apps on one side and being an aspirin or relief to the open compute guys. That's exactly. Is that the best I would, okay. So all right, that brings me to the next point. So your contingent, your growth is based upon the fact that are more people doing more, one, open compute on-premise, bare metal. And two, are they migrating to the cloud? Yes. Two megatrends? Yep, those are to make. And if you go, you talked about Cisco. If you look at what Cisco is announcing these days, application-centric infrastructure, that's all along the same path. So everyone seems to be disrupting Cisco. That's the big whale, right? So I guess, you know, if this was the homebrew club, you would be disrupting Cisco. You know. You know. We did it again. Yes, you did. Yes, if we are top of the rags, which yes, we are competing with somewhat Cisco, but I think the goal here is. But again, the trend is not their friend right now. They're an incumbent. So that is the key part, John. I don't think Cisco would have held on to top of the rack whether I existed or not. That is already going away. That is already becoming open, basically, right? So we are just leveraging the trend to add more capacity and capability to this. Yeah, it's not like you wake up, hey, let's go kill Cisco tomorrow. No, exactly. It's just evolution to you. Exactly. And to them as well. Even Cisco even knows we exist. So that's a good thing for us. Yeah, you're getting some crumbs that are falling off their breakfast plate soon to get to their lunch, right? Then maybe dinner. Well, they are. By the time they figure out that their lunch is being eaten, you're going to be, hopefully, well, and cumulus will be on your well, on your way. Mia asked you a question about other companies like Arista have done some good work. Yeah. Again, funded at the time. Dave, we talked to Jay Shree about this all the time. She got funded during the Great Recession. She did. And she pulled it off. But is the market with her right now? How do you relate to Arista? So Arista, I compare is what OCP is doing right now was opened by Arista in networking space. Because if you look at what Andy and Jay Shree did, basically, they worked the entire ecosystem of the hardware, getting the chip renders to create the right chips, getting the whole merchant-silicon world to open up. Yeah. She's an incredible person, by the way. She's one of our most dynamic cube guests. I can imagine that. She's like her clock speed is, Dave and I were like, we're all amped up. Trying to keep up. We were like, pedal faster. Right. So OCP is picking up some of the stuff they started, basically. Yeah. That is pretty much it. And we are coming more from an operating system angle, building an operating system that understands networking. Okay, let's get down onto the weeds now. Let's get into the nuances. Okay, since we're Palo Alto buddies and days on the East Coast, you got NEA, Menlo Ventures, more Davidow, and Overseas China-based private equity firm. Sure. Do you guys disclose how much you've raised? Yeah, sure. It's a lot of it is public. We've raised about $45 million so far. How many rounds? We have done finish round C about 18 months back. C round. Who was your first investor? So our first investor was Jerry Yang. Jerry Yang did the seed round for us. Then came NEA and NDV in round B and Menlo in round C. Who was on your board from NEA and those guys? Jerry on your board? So Jerry is on the board from round A. Kittu Kalluri is on board from NEA and Mark Sigal is on board from Menlo. Okay, great. Some great investors, obviously NEA is, they don't really tooth their own horn, Dave, but NEA does a good job with exits. They've been really cranking out some serious exits. Obviously they're up there. It's interesting the VC community is changing a lot in the past 10 years, certainly the past seven and five years. A whole new trend has emerged. It's hard to get in this business. Can you share some of your experiences when you guys started? What was it like? Was it easy for you guys when you started? Was did you get a check? Did you have to do the pitches? Did you have some support right away? What was it like in the early days? And talk about that. So we're not a typical startup that way because both me and my partners, Robert Dross and Ken Yang, all three of our very established seasoned guy. So, and we go with the attitude of you got to get something working, solve the hard problem. So we did that on our own dime kind of. And we had already something to see. Then came Jerry about eight, nine months after we started. And the interesting part is Jerry being an entrepreneur himself, immediately recognized and he'd been running a cloud. So he immediately recognized the potential and Jerry was easy. We did in his house, one demo, three hours. I'm being nice. But he's a geek, but he's a Stanford geek as well, alumni. He did not get a very fast decision. You made a fast decision. But he did make a fast decision. By the time the evening was over, he's like, okay, guys, we're doing this. Let's figure out the terms and conditions. We always say yeses are easy, nos are easy. It's those maybe yesterday. Did you bump into him around town? How did you like, you know, cause he's a Palo Alto dad too, I believe, right? Jerry was interesting because my previous experience with Son, we were doing a lot with Yahoo. And that's where I met Jerry basically. Oh, right. Okay, so you had a relationship? Maybe, somewhat, basically. And Stanford grad as well. Then the other part that happened, we came kid to during the NEA, right? And same thing happened. It was during Christmas break. Kid said, hey, everything is light. Come on over, we'll talk. And he was so nice because at the end of the presentation and discussion, he said, guys, don't go anywhere. We're going to do this. You guys are good. This makes sense. So we just need to work out some terms and conditions. We're going to do that. So we got very lucky. Did you do a partner meeting at NEA? What? Did you have to go in front of the whole partnership? Yeah, after, but everything was, you know, well, he was the sponsor, locked him in. Yeah, yeah, yeah, it was the doubt. They're like that. Once you get that partner locked in, like dog on a bone, that's the key, guys. Locked. No, but John, in all honesty, as startups go, because we already solved some of the hard problems before. So things were a lot easier for us. And that's a lot of, you know, Dave and I are entrepreneurs ourselves. And we, you know, we talk about, we funded our own ventures ourselves. And that's the key thing. What people don't understand, doing it early and getting in there and solving that problem, getting some, showing some traction, not just market traction, team traction. And then going to the markets versus the PowerPoint, which some people do get funded if they're a technical team. I know some guys recently, hey, we're going to revolutionize storage. It's going to be up all the way though. They're storage guys themselves. They've done it before. They get a check to start. Just enough rope to hang themselves if they, you know, do it that way. What's next for you? What's happening now? Just give us your personal, what are you watching? What are you playing? I think it's been about four years. Like I said, we are now in stable beta, 19 countries, pretty stable revenue coming in. So two months back, we actually convinced Kumar Shrikantan, who is the GM for Cisco Catalyst line to come over as CEO basically. So Kumar started and his goal is to now scale us, do a public launch and get us. He's going to come in operator like the Frank Slutman, our friend at service now, build out the business. Yeah, yeah, he's a very smart guy. I understand the technology very well, understands the business very well. He had multi-billion dollar revenue for Cisco. So he knows his stuff. I mean, I got to give it to him. He knows his stuff. We just interviewed JR from Rivers from Akumas. They're called the Sons of Cisco. So you have a lot of ex-Sysco guys who have left the mother ship, Dave. Sure. Interesting. Okay, final question, unless you have anything else, you're good? I'm good. Bumper sticker. What's the bumper sticker for this show? What's your take on this? Oh, OCP, what's your, you had to put a bumper sticker on a car. What would it say? Open up the networking. Open up the networking. Yes. This is Open Compute Summit. Great revolution happening. We are excited here. This is Cube all day, wall-to-wall coverage. We're still going to do a couple more segments here. Stay with us. Be right back after this short break. I'm John with Dave here. And Open Compute Summit right back. Thank you guys.