 Live from Palo Alto, in the heart of Silicon Valley, it's theCUBE, covering IO, brought to you by IO. Now here's your host, John Furrier. We are here at the Rosewood for a special CUBE presentation. It's about data centers, it's about the cloud. I'm John Furrier. We're here broadcasting live all day as part of IO Conversations. And we're gonna hear with special guest, Steve Pao, who's the CMO of Ignia Systems. Welcome to theCUBE, good to see you again. Yes, thanks John. So you guys are in stealth, so you can't really talk about what your company's gonna try to get the data out of you. So what are you guys, you guys been in stealth for a while. Talk a little bit about the company, obviously in stealth, so you guys are gonna come out of stealth pretty shortly. But you guys have been in stealth for a while. Give us some history on the company. Sure, so at Ignia's, we're a Seattle-based venture capital funded company. As you mentioned, we've been operating in stealth. We have actually been selling to customers, but we've been really working with the kinds of early adopter customers that wanted to go ahead and take the journey with us. The company was actually founded by a couple of former Iceland guys. So our founding CEO is a guy by the name of Kiran Bageshper. I worked with Kiran 20 years ago when we were back here in Silicon Valley. And Kiran was last the VP of engineering for the Iceland Storage Division of EMC. He, Jeff Hughes, and a guy who named Byron Rikitzes, who was the actually first non-founding employee of NetApp actually all started a company. And what we did, we're in Seattle, we're in Cloud City. We've brought in some DNA from the likes of Amazon and Microsoft. And basically what we're doing is looking to really build a hybrid of systems expertise and cloud native expertise and inform a new company. So what year was that? And so I just want to get specific on stealth timeframe. Sure. One, two years, how many years have you been in stealth? Sure. So the company was actually founded in late 2013. We announced our Series A funding in early 2014. And we have actually been pretty heads down in R&D and working with early adopter customers since then. You know, what's really impressive is the founding team is Rockstars in storage. You mentioned this now, we're in a new world of, you know, it's cloud, it meets systems, it meets software. Obviously NetApp was really, up until I say pure storage has been really one of those storage companies that like Iceland have made it, okay, from startup. And Iceland was huge, impressive success story, obviously, then became part of EMC. But what they were doing big data before was called big data. So you have the chops in storage. You have the expertise on the engineering side, especially with the Iceland, large-scale data. Is that kind of the same itch that he's been scratching, the founding team has been scratching that same itch? And how does that relate to today? Because we all see Facebook, we all see the big names, the web scalers have been called. But now the enterprises are now trying to be like that. So tell me, does that fit into this mission that you guys are going after? Yeah, I think you're figuring it out. So I think that one of the things that's really happened is that that there's been this tremendous data growth and what's really been happening is is this data growth has largely been around machine-generated data where a lot of the analysis is actually done by machines. I mean, humans are basically only involved to actually train machine learning algorithms. And so what we're seeing now is that even everyday enterprises are finding a lot of pressures and dealing with this big data, this big complexity in managing large data volumes. And the hyperscale guys had figured out a way to go do it. And whether it's the likes of, certainly the real hyperscale providers like Google, Amazon, and Microsoft Azure, but even enterprises that are operating at scale, the Apple's and the Facebook's. And what we realized is that the average enterprise can't actually home-grow an infrastructure the way that they've done it, Facebook and Apple and Google and Microsoft. And so how do you go provide that kind of hyperscale leverage to the everyday enterprise? So I gotta ask, I mean, this is the question that comes up and pops in my head when I think about Iceland. In its day, it was cutting edge. But some say it's kind of gotten old and the market has shifted. Obviously with the cloud, with Amazon Web Services, you're seeing massive growth. I think there's storage businesses in the two plus billion dollars that you kind of look at and squint through some of the numbers. And it's continuing to grow where it will be a reinvent kind of unpacking that. But what's different? I mean, let's take Iceland and take NetApp. Solid Fire, so they're trying to modernize. Some would say, hey, Iceland, that's old technology. They bring in the same old Iceland playbook here. No. So I mean, explain, what is the newness of what you guys are doing? I know you can't try to get the data out of you from the launch and you wouldn't try to hold it back in because you have the big launch coming up. But you're a briefing analyst, so what's the new? What's the new thing? Yeah, so I think that really the play here, and this is, all architectures right now are really about how you do distributed systems. And so when it comes down to it, the way that folks like Google or Amazon have built their infrastructure is by loosely coupling distributed systems. And what they do is they build up clouds of lots of identical components. And what that does is that gives you scalability because you've got lots of identical components. But moreover, what it gives you is resiliency. So there's a term, raise cattle, not pets. Have you heard that before? Yeah, I have, yeah. Yeah, absolutely. And so the whole concept is that a pet, you have to do a lot of care and feeding of, but cattle, you can just continue to scale. And if something goes wrong, you can let that cattle go to pass. And that's really the concept that's actually been built in these hyperscale cloud providers with loosely coupled distributed systems. And I think that really the play here with Ignious is to take advantage of that kind of technology. In fact, today we've already gotten 11 patents and issued by the US Patent Office around how we're actually scaling down those loosely coupled distributed systems really for enterprise scale. I'm looking forward to the launch. I can't wait to see what the specifics are. I can't dig into that. But you mentioned the cattle versus the pets analogy. I love that metaphor. I think there was one more. I had someone, Cube, get a say. He brought it to a whole other level, which went more serverless, I think it was. Absolutely. There's a whole other side, you don't like cattle, you can go serverless. But what that points to is the pet vendors were out there, like, oh, I have a server, I'm a rack and stack, top of rack. All these data center enterprise terminology were really pets. Absolutely. And I think that's the real thing. You got to, I love my box. I love the speeds and fees. And you're saying, no, cattle's much more of an aggregate. Yeah, that's right. So that's the real key, that most of these enterprise, traditional enterprise equipment, vendors really design their architectures around highly transactional models. I mean, even an Isilon is based on a highly transactional model. And what we've done is made things a lot more cloud-like. So, you know, the core of our technology is basically about how you take distributed systems and bring that down to an enterprise level. Well, Peter Barris and I have a new podcast called Cube Fridays. Every Friday we talk about the trends and folks watching go to soundcloud.com slash cubecasts or go to iTunes and search for Cube Fridays. But one of the things, and we do that every Friday so you can hear our opinions and certainly it's very opinionated, very colorful. But one of the things we were talking this past Friday on our Cube Friday podcast show was that the data center business is shifting obviously to the cloud. And I think all net new applications, whether it's enterprise or cloud native will be in the cloud in some instance. But the cloud isn't necessarily Amazon web services. It's the enterprise that's kind of wanting to get out of the data center business. So is that kind of what you're seeing with the data? Is that, you know, there's actually some data you can't put in the cloud. Is that what you guys are targeting and talk about that dynamic of net new applications and data moving to the cloud? You know, absolutely. And you know, the cloud has, you know, so many great attributes. I mean, if you're an application developer, you know, the whole thing that's cool about the cloud is that you don't actually worry about the details of how to scale the infrastructure. You don't worry about maintenance. You don't worry about software updates. You don't worry about failure management or troubleshooting. Really, what you do is you interact with the cloud via APIs. And so the cloud is API driven and automated. And what that's done is that's turned IT guys from the buyers and maintainers of stuff, you know, into business partners with their end user groups that really identify requirements and to really ensure that. Architectural kind of conversations. That's right. Architectural kind of conversations and functionality, you know, conversations. And so there's a lot of power to the cloud. I mean, this whole notion of being serverless is something that we really embrace, which is that a developer who interacts with APIs doesn't worry about how many servers are behind the scenes. They don't worry about, you know, auto scaling and up and down. They just interact with APIs. And so that is, I think, the real essence of what's bringing the cloud attributes forward. But I think that the challenge, as you've talked about, is that the data sizes are really growing. There is data that simply, in many cases, can't move to the cloud. And in many cases, organizations don't want it to move to the cloud. So how do you actually get the attributes of cloud but be able to do that at land speed behind your firewall? Steve, I want to ask you a question. And I want you to take a minute to explain, what is serverless mean? I mean, even in the industry for a long time, what is serverless? What does that mean? What's that concept mean? It's got to be server somewhere. But we hear a lot about that. That's right. I mean, behind the scenes of serverless computer, obviously, are servers. But the concept here is that you abstract away the details of the servers through APIs. One of my favorite examples of serverless computing is what happens in AWS Lambda, as an example. Great metaphor, which is that even on launch day, word of ogles actually went in and demoed how you can ingest, for example, images and create thumbnails on the fly. And basically what you do is I view venture-driven computing like modern-day database triggers, where there's an event that happens. You have a small piece of code that fires off when an event happens, like you do a put operation on an image, and all of a sudden you fire off a piece of code to generate a thumbnail. When you write that little piece of code, you don't actually worry about what server it's running on. You don't worry about load balancing. You don't worry about any of the details of scaling that infrastructure. You just write the little piece of code and the infrastructure is transparent. So invisible servers, basically, but there's still servers. You don't have to deal with the configurations. That's right. All right, so you're doing a great job, by the way, holding back all the stealth data. So you got the launch coming up. But I want you to share with the audience some color around what you've been hearing from customers. You've been doing the analyst tour. I know Stu Miniman from Wikibon was briefed. That's right. A lot of other folks, what's the feedback? What's some of the anecdotal or specific commentary coming back to you guys on the reaction, the positioning? Can you share some color around? Yeah, I'm trying to get the data. Yeah, no, I can share some color. I mean, we're not actually going to talk about the specific offering, but I think that what people are basically saying right now is that, hey, the problem that we're solving is a problem that folks are really seeing, that there is actually this set of data that you can't or won't move to the cloud. I mean, obviously, we don't view ourselves competing with the cloud. The cloud plays, I think, an interesting role, even when you choose to have your data on-premises in a lot of different ways. For example, you might want to replicate to the cloud for off-site redundancy. You might want to replicate to the cloud because you need to collaborate. We're working with right now a lot of folks in scientific computing, where the grants are funded by organizations like the NSF, where you're obligated to go to share the results of your data. So the public cloud is great for that. And then the third thing that we actually see is that the cloud is still great for elastic compute. Like, if you're in the media industry and you need to go do a re-rendering job or if you're in scientific computing and you want to apply a new algorithm, there's a one-time job that you may want to actually burst to the cloud. So we're not necessarily competitive with the cloud, but rather a really, really great complementary solution of the cloud, where you can utilize cloud-like metaphors, cloud-like APIs. You can basically have the same tool chain that you're using in the public cloud but do so on-premises. It sounds like a lot like what Oracle's trying to do. You know, there's actually really. For Oracle, for Oracle systems. Yeah, for Oracle. For Oracle, and I think that one of the things is that, and this is going to sound interesting, where really, really big data is not done in relational databases like Oracle. For example, just look at your iPhone. If you look at your iPhone, probably the number one thing is probably your videos or your music or even your text messages, which are unstructured data, because they have embedded images and embedded videos in your text messages. So it turns out that really, really big data is unstructured data. So for example, we're working with a large manufacturer who has sensors, who has sensors in their R&D labs, basically, which generate data on the order of terabytes per hour. And that tees up the whole internet of things trend, as well. Absolutely, and we're seeing that. And so whether it's sensor data, even in media data, we're watching the transition from standard definition to high definition to 4K to 8K. Now you're looking at virtual reality. Huge, huge, huge stage volumes. In scientific computing right now, I had the privilege of being able to see a lattice light sheet microscope and the data that comes through that. Those things generate data at a rate of 800 megabytes per second. So these are huge, huge data volumes. And to do a little bit of writing the data trend and the cloud trend for enterprises. That's right. That's right. Cloud native means systems management means systems definition. That's right. We actually have either coined or borrowed a term called data-centric computing. And so really the problems that we're trying to go solve are really around data-centric computing. And it is in this realm of data-centric computing where traditional IT infrastructure has just gotten way too complex. Like we were working with an industry analyst who cited that in traditional enterprise storage, every petabyte of data that you manage, it turns out there's always a hardware failure. You've got so many disk drives in that kind of infrastructure that it's been cited that you have a full-time head dedicated to every petabyte of storage that you actually want to go manage. And so it's really gotten past this tipping point. Steve, thanks for swinging by our studio here, the Rosewood for our special CUBE event here with IOD centers and sharing your insight. And quickly, when's the launch date? Do you guys have a date? Do you guys know when it's going to come out? Absolutely. So we're going to actually take the covers off of what we're doing on October 11th. So we're definitely looking forward to catching back up at that point in time. Hot new start. Obviously, they've probably got board meetings right here on Sand Hill Road in Menlo Park, we're in the heart of Silicon Valley. I'm John Furrier. You're watching the CUBE, we'll be right back.