 Hey, welcome back everybody. Jeff Frick here with theCUBE. We're at Open Compute Project Summit. It's a mouthful. It's in Santa Clara. It's a 2017 version. It's basically really about the open source spec that was originally part of the Facebook data center spec that they opened up to everybody. And now you've got all the hardware manufacturers, power supplies, fans, everything you can imagine is here because this is really where they build clouds. This is the thing behind the wall. Though somebody else's computer, this is where they're building them. So we're excited to talk to Gary Smerdan. He's a president and CEO of Titlescale. Welcome, Gary. Thank you. It's great to be here. Absolutely. So there's a lot of big companies here, right? But really the innovation, as we all know, comes from startup. You got to start up. So tell us a little bit about Titlescale for the people that aren't familiar with the company. Yeah. So Titlescale, we're doing software to find servers. Now software to find, that's a pretty common term out there, but software to find servers is new and it's important, right? We view it as the final missing piece in the software to find data center. Now the core of our technology is the ability to take one, two, three, five, 10, even a whole rack of servers and get them to operate as one virtual server. So think of it as kind of the inverse of VMware. VMware, you take a server and you cut it in a bunch of small pieces. Right, right. We are the inverse of that. We take one, two, three or more and then to get them to operate as one system. So essentially we can give you the server of tomorrow today by combining kind of standard commodity servers. So what are the applications where I would want to bring them together and operate them as one? Yeah, great question. So there's many applications, but think about big data. In big data, you never have a server that's big enough, right? So we can give you the size of the server that you need. So today what you have to do is essentially you have to take your data and break it into a bunch of pieces. That's humans that have to do that. Spread it over a bunch of machines. We allow the server to be sized to the data. Interesting, so a little bit about the company kind of where are you in your growth? Delivering product, POCs, give us a little background. Yeah, so the company was founded in 2012, really got going in 2013. Dr. Ike Nassie came out of retirement. He had done many things in his career, including being the chief scientist in EVP at SAP when HANA and in memory was all kind of created, and the focus on that. So he founded the company, was able to pull together an amazing team and from there began to work on and build the technology. Last year we began shipping our first product. So we're actively in POCs with partners, with customers, and really looking forward to taking this broadly into the market. So what brought him out of retirement? You said we talked a little bit off camera. He solved a lot of hard problems. Is it just a technological problem? He couldn't sleep at night? Why tackle this really hard problem? Yeah, so basically he had a long career working with folks like Gordon Bell, very famous, the father of the Vax and the mini computer, a lot of other things. Headed up all the software at Apple. And then the experience at SAP kind of put it all together when he retired, he realized, my gosh, there's actually something that can be done that's different and how a server is really conceived. We were talking about it. VMware was one of the final, kind of the last big innovations in server technology. We view that this is the next wave, software defined servers. Yeah, it's interesting because where we hear a lot is the software defined mantra now being applied to networking, right? Software defined networking, because people said, you know, the compute's already been done. There's a lot of work happening on the storage. Now the shift is really on the networking side. But what you're saying is there's a whole another opportunity, another iteration on the compute side to bring a new level of innovation. Sure, and this is all really complimentary. Software defined storage, software defined networking, software defined, you know, it can be as far as the management. All of that is very complimentary. But the final piece that hasn't been, you know, software defined or really mobilized yet is the server itself. And that's the core of our technology. And if you think about it in the modern data center, that's actually the biggest spin is on the servers. So sometimes you need a small one, sometimes you need a big one. And we have the agility and the ability to basically to size it to the requirement of the data. Right, and is the use case where you would be dynamically changing that based on the particular workload or is this something you set and forget and now you've got this SuperMondo server that's dedicated to a particular type of task? Or are you spinning them up and spinning them down depending on what kind of your demand is? So we can do all of those things, but a very typical case is that somebody comes and says, I need this size of a server, right? And so then at that point, you can basically select how many physical servers? One, two, five, and you add up all the memory, all the cores, all the IO. All of that is what is represented up to the operating system and the application. So in that can, at boot up basically in about four or five minutes, you're up and running with the server of your dreams. Wow, and what's kind of the overhead, what's your loss in terms of the management overhead of multiple from as a percentage of the individual some of those parts? Yeah, so it varies a little bit by application, but a typical number when you look at virtualization is people talk about about a 5% kind of overhead associated with virtualization. So that's where we start. In some cases, we are actually able to outperform what a bare metal system would be able to do. And again, when you're able to create a system that where everything can be put in memory, that's how you get high performance in the application. So before I let you go, and it sounds like a great story, I look forward to learning more, but your impressions on the show here, you know, it's interesting up in the city today, you know, SAP's on stage with Diane Greene talking about HANA on OCP, which is pretty amazing. But down here, this is like the nitty gritty. This is, I feel like I'm in the steam room, you know, with Scotty. This is where they're actually building this stuff. It's not abstracted behind the layer, which is what all the customers up in the city are buying. So just your impressions of the show. This is a really exciting show. I was at the very first OCP kickoff that took place at Facebook. So this is a great place where a lot of the, kind of the best and brightest in the industry get together, right? And that includes developers, it includes partners, and you know, everybody come together. So it has a buzz about it, it's great to be here. And we think this is a great place to really be talking about what we're doing with software-defined servers. I actually saw Mark Shuttleworth standing over by it, literally by the water cooler before we went on, Chris Wright on, so you're right. There's a lot of big, smart guys that are here. Yeah, and it's guys like, you know, Canonical with Ubuntu, that runs on top of a software-defined server without any modification. So we run that, we run Red Hat, we run Windows. All of that runs on top of it. And that's a big part of the ease of use that is what really in customers are wanting and needing. All right, well, Gary, thanks for taking a few minutes of your day. We'll be keeping an eye on you and title scale and watch your growth. Great, hey, I appreciate the time. Absolutely. He's Gary Smerd and I'm Jeff Frick. You're watching theCUBE.