 I'm Dave Wright coming in, hello David, thanks for coming in the Cube, and my co-host David Floyer is going to join me here, so we're going to talk a little startup action, flash, disk, tape, what's happening out there, so tell us a little bit about your company here in Stealth, so what kind of signals can you send some smoke signals out to our audience? Yeah, we can talk a little bit about what we're doing, so SolidFire is a next-generation storage company. We're focused on serving the needs of cloud service providers and particularly focused on the challenge of primary storage in the cloud. You know, we've seen the growth of cloud storage for backup and archival, but now we're seeing a lot of applications moving into the cloud completely, people running their virtualized infrastructure completely in cloud service provider infrastructures, and as they do that, there's a need for high-performance primary storage in the cloud, and that's really the challenge we're focused on solving for these cloud providers. So I have to ask you, were you at the brocade party last night, or? I was not. The music was very loud, we were there, and you know, I tried not to yell too much, but my colleague John Furry's voice a little shot, it's dry out here, isn't it? Yeah, it is, I've been talking for two days straight. Good, that means people are interested. Yeah, absolutely. So I've got one question for you, David. That's a tough place to go into, the service providers, isn't it? Why did you choose that place to go after? Why not the enterprise or why not lower down in a stack? It's a good question. So my background's in the service provider space. Where were you before you started your company? I spent some time at Rackspace. Rackspace acquired my last company, which was in the online backup and storage space, and so I spent some time there, really got some good experience building and managing a large-scale cloud environment and saw the challenges, particularly around storage and particularly around primary storage, and saw the need that was there, and really the poor job that the enterprise guys were doing, solving those issues for the service providers. Right. So tell us about the name Solid Fire. What's the implication? So the solid part comes from the architecture of our solution, and it's an architecture built around solid state storage. We think it's probably one of the first architectures that's been built completely from the ground up around solid state, removing all of the legacy spinning disk, and focusing on what you can do with solid state if you start from scratch and build a storage architecture completely around solid state storage. As opposed to sort of bolting it on existing storage array or? Yeah, we're seeing a lot of solid state in the market today, primarily as an add-on to existing storage solutions being used as a tier of storage, as a cash for storage, and you're getting some benefits from that. But I don't think you're really seeing the full benefit of solid state, and you're not able to optimize these solutions around solid state. You've got architectures that have built up over the last 30 years around disk, and augmenting with the solid state is not really solving the entire problem. So can you lift the Kimono a little bit? Well, what's the secret? Show us a little way, Dave. Come on. To optimize this. What makes you think that you can get rid of all these high-spinning disks? So the two kind of really interesting things about what we're doing that we're starting to talk a little bit about is, is number one, the performance piece. And when we talk about performance, obviously using solid state gives us a lot of performance, you know, orders of magnitude more performance than disk solutions. But it's not just about the raw performance. It's about delivering usable performance to thousands of applications in the cloud, and being able to guarantee that level of performance. And that's something that's very difficult to do with traditional storage systems and particularly disk based systems, just because of the physics of the system. And so a lot of what we're doing in architecture is around delivering guaranteed performance to applications in a cloud environment. The other piece of it is around efficiency and about finding ways to make an all solid state system affordable. And not just, you know, make it competitive, but in many cases make it more efficient from both a capital and an operational cost perspective, then spinning dissolution solutions and really making it a no brainer to move to an all solid state system. So how do you how do you do that? So we're not talking a lot about the internals of the technology. But what we are doing a lot of is focusing on efficiency, focusing on reducing the actual footprint of storage that you need to buy and manage and making it significantly more efficient than disk based solutions are today. So just conceptually, I would think that anything you could do to solid state from efficiency standpoint, you could do to spinning disk. So from a cost standpoint that I'd love you to correct me if I'm wrong, but that parity, you know, or disparity will remain. But is the premise that at some point, the performance just becomes so compelling, and you reach some below some cost threshold that it's just a wave of adoption? Well, you know, I think, you know, your first statement about whatever you can do with solid state, you can do with this isn't isn't necessarily true. The physics and the performance levels of solid state are so different. The architecture that we're building around it is so different that that we really are able to do some things that just aren't possible on disk architecture. If we put it on disk, it just wouldn't function. And you certainly wouldn't get the efficiency that we can get with solid state storage. So I think that's a perception that needs to change. But the other thing is, most, most people particularly in primary storage are not struggling with capacity. There's so much capacity available with disk. It's a performance challenge, and it's about optimizing performance. And so it's really kind of flipping the problem on its head with solid state, where you've got tremendous capacity or you got tremendous performance available, and it's about making optimum use of the capacity. And that's really what our technology is focused on. Isn't a big part of the performance challenge of obviously the mechanics of the desk of the biggest, but isn't there a component of it? Maybe you can help quantify this that it's just the overheads associated with, you know, storage protocols, the SCSI stack, etc. Can you talk about that a little bit? Yeah, I mean, I think, you know, you're seeing a lot of move to things like a fusion IO, which are improving latencies through not just using solid state, but actually putting it directly on the bus. And for certain applications, that level of performance is absolutely critical. You know, we're not necessarily trying to replace the entire storage stack. We're just trying to replace and update the part of it that that's 30 years old, the part of it that really hasn't changed in decades, which is the disk. And that's the biggest bottleneck there today. And the rest of it, you know, we've seen tremendous improvements in processing power and controller speeds in networking speeds over the past 30 years. But the disk itself fundamentally hasn't changed. So so the new way of putting this is that previously tape socks and now you're saying disk socks. Is that right? Absolutely. And I don't think we'll find too many arguments with that. You know, disk is great for capacity constrained applications. It's great for backup archival, you know, some big data applications. But anything where you actually want to work with the data you're storing, anything where access is really important performance latency is extremely important. disk does suck and it has for a long time. You know, David, I think you wrote something like that recently, didn't you? Yes, I wrote around architectures of the future. And I said that the central piece of the architecture would be processes and solid state storage in the middle. And then you'd have another layer of storage around it for your archier archiving and low performance stuff, which would be shared. What's interesting about my premise was that a lot of that store that solid state storage would be on the servers themselves. But this is an interesting addition to that architecture, which is essentially a shared storage to the processes, but with much greater performance. Yeah, your discussion, the research note that David wrote was really around FusionIO, the whole concept of, you know, FusionIO is not a storage company, it's a systems company. And that whole thing, one of the things you talked about in that article, and I wonder if we get Dave's sense of this, is Facebook in the data center in Facebook. Of course, that's a big FusionIO customer. And, you know, is Facebook a bell weather or an outlier? Dave, thinking about service providers and the data center design of service providers and how conceptually your solution might fit in. What's changing in terms of data center design, particularly in the service provider space? Sure. You know, so I think that the big change that's coming to service providers is this cloud technology. So we're moving from service providers that we're primarily focused on co-location or managed hosting on dedicated hardware infrastructures. And now we're moving into this cloud hosting infrastructure where you've got much more density. You've got a lot denser architectures with more compute in a smaller space, more customers sharing hardware, sharing storage. And that is increasing the pressure on the bottlenecks in the system. And storage continues to be the big place where their bottleneck and where we're not seeing the dramatic improvements that we've had in compute and networking. That just hasn't been forthcoming in storage. And so that's really where we think we can help break the deadlock, break the gridlock in the data center is by catching the storage piece up to what's going on in compute networking. So talk about where you're at with the company, what can you share with us in terms of funding, headcount, when you're going to be unstealth? Yeah, so we've been in development for a couple of years now. We've got a great team of people on board. We've got an engineering team that's a mix of people from data center backgrounds, from virtualization backgrounds, as well as from traditional storage companies. A significant portion of the engineering team are guys that came from left hand networks and HP. So we've got some great distributed storage DNA as part of the system. The company's well founded. We raised our series A round in January. We've got about 12 million in funding to date. The series A round was led by NEA, as well as Valhalla and Novak Biddle. So who was backing you? NEA is the lead investor. Valhalla and Novak Biddle are invested as well. So we're well funded. We're moving fast and we're looking to get our system and customers hands very very soon. Excellent. Yeah, excellent. So and I'm personally rooting for you because I love going to Colorado and there used to be tons of storage companies. Back in my days at IDC, you know, storage tech. Digital was in Colorado Springs and you know all those optical storage companies out there that you're too young to remember. But you know, storage was a real sort of, Boulder was a real innovation center back in the day for storage and the storage tech loved to see it come back. Yeah, absolutely. There's a lot of great storage DNA in the area and we're looking to to hire a lot of it. Yeah. All right, Dave Wright, Solid Fire, CEO, new stealth company. We just got a glimpse. Tape sucks, storage sucks. Well, sorry. Tape sucks, distuss sucks. Storage doesn't suck as we're here at EMC World. We can't say that. Throw us out. But thanks very much for coming on the queue. Thanks a lot. I appreciate it.