 Welcome back. We are here at the OpenStack Summit in Portland, Oregon. It's a beautiful day outside. You're in the Cube, SiliconANGLE's exclusive video production, where we go out to the events, we extract the signal from the noise, we really take you into the events with the key players that are here to give you the inside story. And we're really happy to be joined for this segment by David Floyer, our Wikibon analyst extraordinaire who's sitting in for John, which we're happy. And we have Dave Cahill, who's the Senior Strategic Alliances for SolidFire. Sorry, it's been a long day already. So I'm just going to kick it over to Dave, and why don't you give us a quick update on SolidFire and what you guys are trying to get a couple of key announcements today. Yeah, sure. So SolidFire is an all SSD storage system targeted exclusively at large scale, multi-tenant cloud infrastructure use case. And obviously, prevalent in that discussion as any is OpenStack, because all of our customers today, when they have a conversation about deploying, managing, running a cloud, OpenStack's at the center of that. So that's a key integration for us. We have invested in the OpenStack community from very early on. And in fact, we're integral in helping extract the block storage projects that is now Cinder out of the NOVA project. Today, we announced some advanced developments there. We continue to lead the way in terms of the industry's broadest support for Cinder, ensuring that we have compatibility with all of the Cinder-related features, including boot from volumes, QOS, multi-instance back end, but also added beyond that as well. So we're also driving meaningful ecosystem partnerships with both Nebula and Rackspace private cloud. And also starting to show off some of our early customer momentum with customers like Brinkster, who are now deploying SolidFire behind their OpenStack-based cloud infrastructure. So trying to pull together more than just technology support, but support for the community, support for our partners, and also now customer momentum as well. Great. So the SolidFire is obviously an all-flash array. And what's the thinking behind you supporting the Cinder as opposed to others in this area, the more traditional storage arrays? What is it that drove you to do it? What's important to the community for you doing this? Yeah, so I think our support for OpenStack and for Cinder really is a function of our focus on a specific customer set. And that customer set today is large-scale service providers, cloud providers, public cloud infrastructure as a service. And those guys are looking at a few different building blocks behind their infrastructure. One of them is absolutely open-stack. The others are probably cloud-stack or something like VMware. And their goal is to deliver an integrated solution in support of particular cloud services. We need to be relevant in the open-stack conversation. We need to be integrated out of the box. And also SolidFire has a lot of advanced functionality that we want to ensure is accessible natively through Cinder as opposed to forcing a second or third integration point. Our goal is to ensure that storage can be dynamically adjusted, managed, provisioned just like network and just like compute. And in order to do that, you have to drive API level integrations with the key frameworks that exist. And OpenStack is absolutely at the top of that list. And you can see 3,000 people here this week that it's only, the momentum is only picking up. Absolutely. And congratulations on picking a winner like this early on. That's tough to do. But what sort of applications are driving the need for low latency? Why not just a simple traditional disk drive array that's out there and well understood? Yeah. What things are you introducing that make it really different? So super low latency isn't necessarily the use case that we really care about most. We're not going after the one or two percent of applications that need performance at any cost, right? We want to deliver consistent predictable performance to thousands of applications in parallel. Where is that most prevalent, the cloud infrastructure use case? And that's where we see guys today demanding levels of quality of service. Once you can deliver QoS on a per application basis, you can bring that use case to life. You can host thousands of applications in parallel, essentially isolating them into their virtual resource silo. They run next to each other unencumbered by the noisy neighbor. And that's the key use case for us. Why can't disk do that? Because it's more performance constrained. Our goal wasn't to deliver a really big footprint of IOPS or a really big footprint of capacity. It was let's deliver the optimal balance of both. And then so let's deliver the technology that allows the customer to provision as they want across either depending on the needs of the application and not for some unnatural optimization based on the limitations of the infrastructure. Excellent. Excellent. Okay. So how's it going for solid fire? You've been a little later than you would hope to in terms of getting out your initial products, etc. How's it going? How's the business going? Things are great. So we had an early access program that extended about a little bit longer than a year. We announced GA in November of last year, with four service providers as customers. We sense follow on from that momentum at a strong end of the year and early good start to Q1 as well continuing to drive more points on the board in terms of meaningful customer deployments from large scale service providers. Also seeing some really genuine interest from large enterprises that have an enterprise logo on their front door. But frankly, under the covers look feel and run their business like they're a service provider. So so do you have any success stories with open stack and Cinder that you've got in beta at the moment or you can talk about? Yes. So one we actually included in our announcement today is Brinkster. Infrastructure is a service provider out of Arizona, who very much desired the need to deliver quality of service to their customers, eliminate the noisy neighbor problem and evolve the complexion of their applications that they could host in their cloud infrastructure. And so you know, for us, the reason customers seek out and choose solid fire is very much around business enablement, evolving from hosting our managed services business to to take the flag in this cloud opportunity. And they've used solid fire as a bit of an arms dealer in that regard to help them capture that flag. And so Brinkster, I think has looked at a lot of different technologies from a cloud perspective and settled on solid fire. But above that, more importantly, settled on open stack. And a key piece of that was that, you know, we have the industry's most comprehensive support for Cinder. We have strong institutional knowledge for Cinder, and just have been around and involved in the open stack community for quite some time. Dave, tell us a little bit about the economics behind it, because I think there's a pretty popular and maybe it's a misperception that drive arrays and kind of traditional drives are cheaper. Yeah. And if you're saying that that that the driver of your business is not the super super low latency, super high performance, it's more consistency of delivery of service. How does that work in from an economic model to justify folks going with with that configuration? Yeah. Yeah. So I think there's a few pieces to that. The first is, you know, you have to be able to crack the code on the economics of flash. And you know, there's a continuum of mediums for different types of information. And, you know, if you can crack the code on the economics of flash and open it up to a broader use case, you know, that's that's part one. And how do you do that? You can compress, you can dedupe, you can thin provision the data set to drive a much larger, larger, effective capacity footprint versus relative raw capacity. Okay. And that's, you know, that is really important to open it up to a broader use cases, driving down the effective cost per gig of flash. And this isn't a new story. We saw this with data domain. Once data domain cracked the code on the economics of disk for an archive medium, all of a sudden you push tape library to the right and insert a data domain into that discussion, a disk medium for archiving. We have customers today that are using all SSD platform for all of their cloud solid fire. What if they don't want their disk, they pushed it to the right and turned it into an object store. That that trend will only continue to work in our favor as the economics of flash continue to go down. That's, you know, economic conversation one. Part two is around application density. If you can deliver a large bucket of IOPS that can host thousands of applications, and then you can apply qos to every single one of those applications, then your storage cost per application is dramatically lower. It's no longer a one to one storage to application ratio. Now it's one system and the cost of that system is amortized over thousands of applications. Your per application cost for storage is significantly lower. That's the moment where guys start to get excited about the fact that this is not a there's no longer that rigid binding between application and storage. When you can prosper from shared economics in a multi tenant infrastructure. That's when guys forget about the fact you're talking about flash. They're talking about business enablement and economic prosperity for them. Because you've really changed the conversation. Once you've broken that one to one relationship, now it's a totally different kind of conversation. Once you get them over the fact that flash is not just performance at any cost, then they're ready to have the conversation about the operational benefits of qos. So you have to kind of normalize before you can operationalize. But that sequence is something that people are getting more and more accustomed to. I was going to say how do you see the market kind of recognizing that that part of the story and getting away from is just your performance. You know, so I can only we can only control so much of that in evangelism. We fully understand that we have to be out in front of evangelizing the fact that flash is for more than just the performance at any cost use case. We know we have to evangelize around the types of applications you can host once you drive qos. But part of it's also just time familiarity. I mean, this conference two years ago was 500 people 200 people on the second floor of the intercontinental in Boston. Now it's 3000 people in a convention center, right? Time has allowed people to get more familiar with OpenStack and the use cases and how to use it. The same exact thing is going to happen with flash, right? And the use cases and broadening of that and the opportunity for us will follow. Great. So just asking you, it's moving very fast. Where do you think you are going to be as a company and as part of OpenStack in one year, two years time? What's the outlook for yourselves and for Cynder and for OpenStack? So, you know, I think that the community momentum around OpenStack is phenomenal. You know, we are going to continue to invest in this community. There's nothing stopping us. I mean, we it's institutional for us, right? Our founder came out of rock space. He understands OpenStack. I think you're seeing there's only there's limited options here right now. You're seeing some noise starts to sort away and it's mostly signal and you have OpenStack against CloudStack against VMware. In some cases, you know, we're Switzerland and that we want all of them to do well because that's what we want. Whatever our customers want to do well from a stack perspective, frankly, we want to be able to ensure that all of the advanced features we're developing are populated up through OpenStack and we can take advantage of those. So, you will not see us at all decelerate our investment in OpenStack. Will the market vote away from OpenStack towards something else? I think that's to be determined. Where I get nervous is if this evolves away from a community and more towards a standards body. And the bigger the vendors that kind of come in tops down, the more nervous I get. So, the community's ability to hold those guys at bay. I don't think, you know, you can throw money at this problem. Really, we want contributions, right? We want contributors, developers, people evangelizing the story. I don't think OpenStack has a real problem with money right now. But to the degree that people are just kind of waving their hands and glomming on and tying their wagon to this horse, that's when it starts to dilute the community effort. And if the big guys dilute the community, then that's a larger issue. But we're going to watch it closely. And I think actually, right now, the complexion of the conversations at this conference are changing completely. Guys are now thinking about deploying. They're not necessarily super enthused with VMware's cloud strategy. And they want to look at alternatives. So, OpenStack is as good a shot as any to take that flag. Yeah, it's interesting because that's been a pretty consistent theme of the people that we've had on the Cube over the last couple days, as people are contributing code and they're raising their hands. We want to contribute code. We want to be part of the code. We got into it. We had our last speaker, John from Raxby, saying, we had guys that are in other open source communities and just kind of love the vibe, love being part of something that's bigger than the company. So we can keep the really smart guys, give them something that's evangelical and really groundbreaking to work on. And it seems to be doing pretty well. But it's interesting that you said, good proper SWAT analysis. What are the threats to this thing going forward? And what are the potential weaknesses? That's a question I ask a lot of people. How many contributors do you have? So let me ask you, how many contributors do you have contributing to the open source? We have one. Just one. And he has by himself, you can look by any contribution measurement you have. John is a phenomenal contribution. But that's all you need. I think what I don't, what you hear is, hey, we're all in. We've invested a ton of money. We're hiring tons of guys. Wrong approach. Get one guy and get him to start doing some bug fixes. That's it. Just start contributing. Get him to go attend a meetup. Figure this thing out, right? Don't do this tops down crap where you just say we're here and we're present. We're more stake, less sizzle. And stake is showing up, starting to fix bugs, taking on some grunt work, proving out your currency that you're here and you're legitimately looking to contribute versus just put out a press release and say, hey, I'm involved and glom onto the what is what, you know, increasingly obvious momentum that's going on here. Separating those two things out I think is important. And we need more of the more of the stake. Well, that's the sizzle. All right. Well, great. Well, thanks for for stopping by. Yeah. David Cahill from solid fire, exciting stuff and in solid state memory and and really transforming the way that those technologies applied to applications in the data center to really change the economics for what I think a lot of us at least people like me have really thought about using that type of technology and it really being kind of out of reach or still a little bit too expensive. So really changing the game to be able to bring it into the house. So again, we are we're here. You're in the cube. We're at OpenStack Summit. Great vibe. We got a full day for you. We're going to take a short break and be right back.