 This is theCUBE, live from the Moscone Center in San Francisco. This is SiliconANGLE's continuous coverage of VMworld 2010. Now, Inside the CUBE. SiliconANGLE.tv's continuous coverage of VMworld 2010 here in theCUBE. We're broadcasting live on the scene at Moscone South in San Francisco, where VMworld is taking shape. We're in our third day, but really the second biggest day with my host, Dave Vellante. Dave, welcome back. Thank you, John. Yeah, day three. Day two for huge, right? Excellent, excellent to be here, John. And in theCUBE right now is the CEO of a hot startup growing like crazy, David Flynn from Fusion IO. Welcome to theCUBE. Great to have you here, David. Thanks, David. I understand you had the invite to the Apple announcement today. You sent WasNSTED. Our chief scientist, Steve Wasniak, had two VIP tickets. So he said, hey, you want to come with me? We get front row seats. Maybe a chance to sit and talk with jobs afterwards. I'm sorry, I have to do a TV interview thing. So, my business partner work is there instead of me. So everyone on Twitter want to know if the Was was showing them it. No, we could have had backstage exclusive access to Apple, but we, doing theCUBE instead, thanks. The big thing right now, we just heard from some of the entrepreneurs out there is pick a market, focus, be disruptive. You're in that right now. Talk about what's going on in your world right now with Fusion IO. You're emerging, you got some technology. Just give us a high level overview of that and where the cloud all fits in and VMware's relationship and all that. So there's a huge disruption going on in the data center today with the advent of a new type of memory that has 20 times the density of RAM and 100 times the density by the time you pack it on a module because it doesn't require power and doesn't give off a lot of heat. Of course, I'm talking about NAND flash and this is hugely disruptive because you have to say what would life be like if you didn't have constraints in the amount of capacity of memory and didn't have constraints in the amount of IO you have. So it's a really interesting market to be in. For the folks out there who don't understand all the nuances of storage, I mean, you're talking about high-end stuff, but they have cameras, like flash cards, right? That means they put it in, it works. That's right, it's the same technology, the same chips that go into smartphones and MP3 players and cameras. But what we're seeing now is that that's being deployed in the enterprise. There's two basic methodologies. One is to make it a faster disk and have it pretend to be a disk drive for a storage and the other, which is the route we have kind of pioneered, which was to make it look like a memory device. As a very high-density memory, it's extremely valuable. So we've been covering here at silkenangle.com.tv at the VMworld, the convergence, the storage, compute, networks all converging. Storage has changed. I mean, talk about specifically how you see the cloud and storage changing, in particular, the role of storage. I mean, it's one big computer. I mean, what's your vision? How would you describe that to folks? Well, it's kind of like in the computer graphics industry. There was a day when you had to have a vertically integrated platform with vendor lock from like SGI, right? They could charge $100, even $200,000 for a single user workstation, right? That's the storage of today. With the advent of the graphics card, it allowed anybody to build out of commodity parts the stuff necessary. So what we're seeing is the storage industry being transformed like the graphics industry was and that the agent, the catalyst behind that is the scale out business model of cloud and the high density of flash, the commoditization of the performance layer that used to take so much of an investment. So talk a little bit about that. I mean, you mentioned the scale out. Now these companies like Twitter and Facebook and the so-called Web 2.0 guys, when they, they grow like crazy and they want essentially building blocks that they can use to scale out. Are you seeing that in your customers? That's right, and you've already listed some of our largest customers, right? With this technologies, those guys, instead of getting so many users on a server, they're able to get 10 times the throughput. So the benefit of having a very high density memory storage hybrid is that those databases get 10 times the throughput and their response times are significantly faster. So the scale out world is being transformed through this new type of memory. And now Wikibon has talked to a number of these customers and they have essentially said that they were, prior to solutions like yours, they were just throwing hardware at the problem and it didn't solve the problem, did it? Well, and the reason is because the servers were really just cheap ways to package spindles or package a quanta of RAM, right? But the CPUs largely went underutilized to the tune of maybe only 10% of the processor being used, the other 90% sitting idle. So many of these properties bought two socket motherboards and only put one CPU socket, one CPU and one socket left the other one empty because if you're not even getting the full utilization of one, why put in two CPUs? It's just cheaper to package your memory in a commodity server package or to package 15 high performance spindles in a commodity server package than to centralize all of that into a big storage array. That's the whole thought of scale out. What we do is we allow them to fit in the server enough capacity and performance of this IO memory that they can get full utilization of those processors. So we've been hearing all week, David and John about cloud scale and scale out. I mean, I've been following this market for years and solid state disk has been around for a long time. Forever, since the dawn of computing. And so we've now got this flash coming in from the consumer market, and so what else is different now? So what's different and the reason why this is relevant today is isn't because flash can make a faster disk drive. It, you know, it always could. You can make a faster disk drive, putting RAM with a battery in a disk drive package. What's interesting is that flash has 20 times the bit density per square inch of RAM. And because it doesn't require power, a hundred times the density on a module. So it's not flash as a faster disk drive that's made solid state all that important. It's the fact that flash is a higher density memory device that has really made for a big change in the air. So translation meaning it's really now cost effective. It's very cost effective. It costs less than RAM per bit and you can pack a hundred times the density on one module and get 800 gigs on one card up to five terabytes. And it's persistent. Now you're right. You could have created battery backups before but it's a lot cheaper to do it with flash. Yeah, and even if it were not persistent, even if it forgot the data just as a memory device it would be huge. But the added bonus of the fact that it remembers the data when the power's off is extremely powerful because now you don't have to wait for all that time for the data to load in. I mean some of these scale out properties like you were talking about have to stage a database and run traffic against it for several hours before it's ready to go live because they have to warm up all of the cash in the RAM with Fusion IO, IO memory holding the data sets. You can go live and it's instantly ready because the data has been persisted there. So what does that mean for an application developer to say developing cloud applications? So it's a totally different mentality because things that you thought were impossible to do are now possible. You have to ask yourself what would I do if I had a system with terabytes of memory and didn't have to wait for that to get restored after a power outage but it was holding it all the time. The fact that this is a hybrid memory slash storage device makes for some really interesting designs. You can start combing through massive amounts of data in exhaustive ways that would not have been feasible before. David, can you talk about some proof points around the applications? I mean honestly it's like having RAM and a PC back in the old days, it's all on demand there. What are you seeing from your business and that you're disrupting and some proof points around? There's new things. It's an interesting question because it spans such a large spectrum because we're talking about a fundamental new building block, right? So it impacts and will impact everything in the entire data center. But we can take out some for key examples. In the database world, like we were talking about, it typically means that a database server can do about 10 times the throughput for the same server. And those responses, the queries, are answered 30, 40% faster. So it means faster page loads, more throughput per server. So answers.com retrofitted their MySQL scale out database tier and saw nine times the throughput per server. What they chose to do was to shrink the database farm four to one. So they got a 75% consolidation and with that remaining one out of every four servers, they were still getting in total more than twice the throughput they had before. So they re-architected essentially their database platform. And one way to think of what we do is we sell server consolidation just like VMware, but our server, we do that on the data intensive end of the spectrum where it requires such a large data set or such fast access that it wouldn't all fit in RAM, it wouldn't all fit in spindles. Well, we cover a lot about data on SiliconANGLE and Dave, researchers with Wikibon. You know, I was talking about big data, Hadoops of the worlds, but there's little data. We've been teasing out this notion of little data. Low latency, fast response times, edge devices. With virtualization, little data becomes big data because you can pack as many workloads as you want. Exactly, you got fast big data. At that point because now you've jumbled all the IO together. So you're fast big data. That's right. That's the way to think about it. Fast big data and in combination with VMware, what it allows you to do is to further increase your consolidation ratios. Because what limits the number of virtual servers or virtual desktops you can fit on a single machine is the memory capacity required and the IO performance required. This fits both of those bills and allows you to completely remove those as constraints. So talk about what that would mean, just playing it forward. Just kind of, you know, the fantasy, you know, in a couple years. I mean, I've never talked about VDI virtual desktops in the end user experience, but people have iPhones and iPads. So what is that experience going to be? I mean, spectrum is a problem for wireless and everyone wants faster everything now. So to explain that enablement, what would be some of the things you'd see? So what we're seeing is that flash is transforming your endpoint devices, right? So you're carrying around even more intelligent devices with more capabilities thanks to that very cheap capacity in the NAND flash and ever increasing processing performance. Where we play is in the back-end infrastructure, what it takes to support those, to run those services in the sky, this makes them much more cost-effective to scale out. So it would be safe to say that, you know, my iPhone will be more powerful, my iPad will be more powerful? Those will be more powerful and they will appear more, you know, infinitely powerful because they're going to have fast access to large arrays of servers outfitted with IO memory that lets them hold huge amounts. So you're like a little memory cache that sits on the edge of the cloud that serves up the edge device, right? Almost a way to think about it. Well, metaphorically, it sits in the cloud and it provides a new tier of memory. So it doesn't displace disk drives, you'll still use disk drives for the cheap capacity that it gives you, but for your active data sets, those will all be stored in IO memory and you'll still use RAM, but you won't have to put as much RAM per server because now it can access the flash-based IO memory. So it's a new type of server class memory really? It's a new type of server class memory. That is the exact way to put it. The fact that it happens to be non-volatile and that it happens to be accessed through the storage interfaces is really a convenience. Just like a RAM disk sets aside part of your memory to pretend to be a disk drive, this, actually it's our software stack called the virtual storage layer, takes IO memory and expresses it as if it were a storage device, but the way it's physically connected, the way it's communicated with is much more akin to how you communicate with memory. So John mentioned VDI before, can you share with us any metrics or again more proof points around VDI and what you're seeing? Yeah, we set up for the show here a demo where we have 512 VDI desktops being served from a single server and the amazing thing is that it's CPU bound, it's a 48 core server and we have in essence removed the bottlenecks of the amount of memory required and the amount of IO. So you can even, you should stop by the booth and see it. You can log in through an iPad and start interacting with any one of those desktops. They're going through a scripted procedure showing users very actively using the desktops. So we've talked about it, but I wonder if you could summarize sort of the difference between the Fusion IO approach very close to the CPU versus taking SSD and putting it into a disk controller based subsystem. That's right. Two general schools of thought, one is to have it pretend to be a disk drive and sit inside of the disk infrastructure, which seems like an easy target for very quick deployment. That would be kind of like having flash chips. You know how you carry a USB thumb drive? Sure. Well see, flash already displaced one form of magnetic media, the floppy drive. Notice it didn't replace the floppy drive by looking like a floppy disk and going in all those floppy disk readers. There was actually a company that did that and they didn't last very long because the USB port was just way too convenient to get more capabilities. Well, we thought the same thing. Flash is a memory device and should be integrated into systems as a memory device to be able to truly exploit the additional capabilities. Because the last thing you want to do with it's more expensive than disk drives. So you don't want to handicap its potential benefit. So we said we should be focusing on how to optimize to extract the most benefit from the flash, not optimize to have it pretend to be like the old school stuff. Right. Can you talk a little bit about how you could potentially share that resource? Is that something that's on the horizon? Well, you know, the thing is is all storage starts local and gets shared out over iSCSI or fiber channel through some sort of a director or service and you can take IO memory and serve it out iSCSI or put Microsoft or open source on it and have it out of filer. So anything that can serve storage over a network can serve IO memory virtual storage out over a network as well. So we're demonstrating iSCSI based targets, et cetera downstairs. Now it's best used locally because of the very high performance. We have cards that have the performance of 16 FC four ports. So just one card and it gets very expensive to try to extract that much. So you'll see a tendency and it's very in line with cloud scale out to have the storage local and have it used locally even if you have a sharing layer of software. For example, Datacore, one of our partners is here showing how with their software they can create distributed shared sand-like storage from servers that have IO memory under VMware ESS. Yeah, we had George Tashara on the show on Monday, I think, yeah, it was good. John, I wonder if we could talk a little bit about David's business, about Fusion IO, the company. Yeah, I mean, what's going on with the status, what's the momentum you have and revenues? As, pause, I think. You bit? As you might expect, when you're introducing a new kind of building block that's this radically different the question becomes, well, how do you drive adoption and make sure that there's traction with it? And I'm happy to say that we have OEM wins with all three of the major server vendors. IBM sells our product under the name High IOPS, PCI Express. It's actually designed into their web sphere appliance, their web server appliance, and designed into their info sphere appliance, their database appliance, already built in. And HP also has OEM the product. Again, sales it as if it were their own. And it's called the IO accelerator. And that's across their server line. And most recently, just a few weeks ago, we announced an OEM relationship with Dell, who is actually sharing our branding on the product. So it's a Fusion IO branded product, but it's a full OEM relationship with Dell. Dell also happens to be an investor in the company from our seed round investor. So you're going to run the table on the server vendors? I mean, Dave and I were very excited that you were coming on, by the way, because we think your company's really hot in terms of like those huge success. And IO is a problem, right? We all want the faster edge device. We figured that we had better work well with the server vendors, because what we're selling actually ends up being server consolidation, right? And just like VMware had some of that constructive conflict with, what does it mean to be consolidating servers? We have that same thing. So we felt it was very important that our products were sold by the server vendors and allowed them to make a business. It gives them more performance. I mean, they can still serve their customers and deliver better value. And you know, the amazing thing is, is one might, at first blush, say you're going to consolidate servers. That decreases the server spend. But in actuality, when you lower the cost per unit of useful computing, you make it more cost effective for problems that weren't before. And now you can take business models to market that wouldn't have been viable. So it actually expands the overall IT system. This is something that David Floyer from WikimHunt has talked about a lot, that compute and storage resources. It's an elastic market, right? It's very elastic. Lower the price, people will buy more. They'll push applications harder. So we may talk about consolidation, just like VMware, but what it really ultimately means is expanding the overall market. So you guys are obviously growing very quickly. Can you talk about headcount, revenues? I mean, or even just give us some general guidance. You know, the company is doing extremely well. Was is there. Everyone wants to know about Was. Steve Wasniak joined the company now going on two years and is extremely active with the company. We're very pleased to have him. He's, aside from being a celebrity personality, I mean, dancing with the stars. I don't know how he did that. Yeah, but actually that happened just as he was joining up. You know, he accepted. We thought we'd go out on a lark and ask him not just to, you know, be an advisor to the company, but maybe actually take a paycheck and be an employee and come in on a regular basis and be part of the executive team. And he accepted because he was so excited. Solid state has always been, you know, special place in his heart. But then he said, but I'm going to be unavailable for the next three months because I'm going to be sequestered down in Southern California for this thing called dancing with the stars. It's like, it worked out really well because now, you know, even people who don't recognize him for being the inventor of the Apple computer, they recognize him for being that guy on dancing with the stars. Good, so uptake is strong. Yes, it's one of those up to the right charts. It sounds like what you guys. The impact that it has on business is incredible. We have some customers who are using us under autonomy and Microsoft's fast unstructured text search. That's, you know, a growing area is corporate kind of search discovery. And, you know, their improvements are incredible. They went from minute level response times to comb through the inventory of every last, you know, email, what have you, down to fractions of a second in the response time. So it can really change the business and the services that they provide. So we're seeing a very rapid uptake. So the reason I think, this is my personal opinion that companies like Fusion IO are so interesting is because not only are you attacking the efficiency play, consolidation, right? It's a big theme, VMware, big theme. But more importantly, and this is true, I think for VMware as well, helping IT become a value producer, right? Whether it's a revenue generator or a new business enabler. And from your standpoint, it's enabling orders of magnitude greater application performance potentially. That's a good way to look at it. And we kind of segment it too. There's folks who want to use our product as a way to reduce cost and improve efficiency. The more forward looking guys are actually the ones who look at the creative ways of, you know, now that this possible, now that this is possible and cost effective, what kinds of things can I do that I couldn't before? And it means thinking outside the box. Just more tactically, what worries you as a CEO? What keeps you up at night? My biggest worries nowadays are around supply chain and building stuff fast enough. It's a problem that we always said, hey, that'd be great to have. Well, it's still just a problem. What do you say, just a final question to wrap it up? It is advice to other entrepreneurs. I mean, you've built a great growing business, great wins. What do you tell other entrepreneurs out there? Boy, run with your dream. Go through, analyze it well, make sure you're targeting a good market and just trust your gut on stuff and be a little bit of a non-conformist. I think that's what's really played well with us. As we said, look, everybody else is trying to conform to those same drive bays and storage controllers. And we said, that all needs to be rethought. So being ambitious enough to take on the daunting task. Okay, David Flynn, the CEO of Fusion IO, hot startup, exclusive coverage of VMworld 2010 Live, SiliconANGLE.com. Thanks so much, we're here in theCUBE. Feel free to mingle with the bloggers, and I'm sure they have a lot of questions. App developers really interested probably in all this. So thanks so much, Dave, great. Great to have you on, David, appreciate it. We'll be right back. Thank you. Taking a break, we're going to have an analyst come on from wikibon.org and Juniper Networks. Talk about networking and possibly some data and security. And computing.