 So now, huge consumer volumes bringing the price of persistent flash down and now that's migrating into the enterprise, somewhat different technology, but fundamentally it's the same, right? So we call that Moore's Law of Storage, okay, so you've known Moore's Law of Compute, where Intel is doing a great job in doubling the CPU horsepower every year. Every year the CPUs come out faster, server utilizations are around 10-12% in these data centers. Yes, flash sustainable. We think this is the way in which we can actually start shortening that gap. And this is the first time we've started using flash in enterprise settings, driving application performance through the roof. It's unprecedented in terms of what that is and how that can be deployed across large-scale clusters is what we are focusing on. So yeah, so I mean... Is it sustainable? Is flash sustainable? Absolutely. From a technology and engineering perspective. So flash itself, and if you see that what ships today is what we call 34 to 40 nanometer flash, okay? This is the lithography that people make these factories, you know, Samsung and Toshiba and Micron make these. And these are going down, there's at least three more generations. They're going to cross over in price points compared to hard drives over the next three to five years. So you can deliver, one of these flash drives can deliver 2,000... So for instance, FC drives, right? All right, fiber channel drives. Correct. Not SATA. No, it's coming down. It's coming down. It's coming down faster than you think. Within five years? Within five years. So within five years, just make sure I get the prediction right. Within five years you're saying that from a cost per bit standpoint... Cost per byte, cost per bit. Will be cheaper than spinning disk, fiber channel and including SATA. Correct. Well, yes, yes. That's my prediction. That's the first time I've heard that prediction applied to SATA. So that's... Because it's getting to a point. An MLC flash drive was under a dollar per gigabyte today and magnetic storage is around 10, 20 cents per gigabyte. So it's a factor of five to 10x a bit. And flash prices for a while were coming down faster than... 60% per year for the last four years. Yeah, but then there was an uptick, right? Because memory price is uptick. But they're moving... So if you look at the next generation of storage, they're moving to three bits per cell. That's going to come down. People are pushing it in different ways. So five years... Okay, so... Imagine the next iPad to iPhone... We're drafting behind the consumer ecosystem. Yeah, so David Fleury... We were talking to David Fleury before. He made the prediction that flash drives, enterprise flash drives, would cross over fiber channel drives by 2012. I think within another three to four years behind that... Pretty aggressive, right? And then prices upticked. This was... He made this prediction in like 2008, right? I mean, that was pretty bold to say that back then. Prices upticked and so it sort of had to revise the forecast there. But so, okay. So you're the first I've heard say it's going to cross over SATA and give a definitive timeframe there. That's huge, right? And his other point was it doesn't have to be cheaper in order to cross over because of the incremental value that you get out of flash. So let's talk about that a little bit. So I want to back up a bit. I've been watching storage trends for a long time and I've seen function go out of the host into the sand, right? Just hanging it back up there. So it can be shared and also the one of those seven. And now it's coming back up. Now it's going to the other side of the channel. But the difference is now this time it's persistent. That's a huge game-changing phenomenon, isn't it? Talk about why. So there's two or three things that's happening here. So one is workloads and data is growing nonstop, okay? So people are... Data is getting digitized. People are doing more with data. They're doing more intensely of uses with these data. So servers are getting faster, data is growing. The natural, natural point is really how do we access it? How do we deliver it? How do you make this efficient from a data center standpoint, right? Servers is where the workhorse is. Today you have to put storage outside. You have to put a thousand drives to get 100,000 I-Obs or more. 10 years ago, 20 years ago. One of our cards today delivers half a million I-Obs today. One of our cards. That's equivalent of 4,000 hard drives and one little card, six inch by three inch. We can put four of them in a server, put 32 servers in a rack. We'll scale this up very, very, very fast, okay? That's the exciting thing. PCI storage is a tenth of the latency, 10 times the bandwidth. Intel is integrating PCI into its CPUs over the long term. Generation three is coming. This is where the action is. This is a new generation. For solid state PCI is the right answer, I believe. Now, we've been talking about Fusion I-O. One of the things Fusion I-O does is they, through software, essentially eliminated storage protocols, right? And so that's a pretty interesting dynamic. How are you different from what Fusion I-O is doing and why will you be able to compete? So, a bunch of couple of simple things, right? What people buy these for is really what we call application sustained performance. You buy these cards, flash is still a new media. People are learning how to do with it. What happens to this flash drive on day one, what is day 1,000 is very different. You have to understand the physics of flash, how this media works, and how do you manage that performance profile across time, across workloads, and across capacity. What we do very well is we are four to five times the performance of the best in class, Fusion I-O today in the PCI sector. We can deliver the performance across a range of workloads, from read mostly to read write bias workloads, and we can deliver it when the card is full and it's utilized to its capacity. So, all three factors combined, we can deliver a value proposition that's a factor that's compelling to the enterprise. Okay, so let's talk proof points and customers. Where are you at with customers and you're showing out some pretty big numbers? That's impressive. We've got great technology. We've got the best of great technology in terms of how... We've done hard computer science to make this work. We have a bunch of algorithms, a bunch of technology in our boards that helps us scale. Proof points-wise, we have launched this product about six months ago into the market. We're building our revenue. We've got over a hundred plus customers who are building up. We are working to get aggressively new channels to the market. We're going to announce a couple of large deals. You're going to see that in the press over the next couple of quarters. Did you talk about any customers or...? I have a bunch of case studies. We can go to our website. Some of the largest... So, we have a customer called iParadimes in the Web 2.0 world. They were an initial Fusion IO customer. You've got very good performance numbers. See, they switched. They switched. Why? Fundamentally, what we can deliver is a sustained performance at a capacity utilization that's unmatched. When you fill out the dry, when you keep executing on that, fill data, we can deliver an order of magnitude better performance. That's the facts. And there's case studies on our website. We'd be happy to give you more pointers to that. So, we've got a couple of other customers who are doing new solutions with these. You know, it's not... You know, this is the first of nine innings in this game. It's a huge market. We've got database solutions being created, virtualization solutions being created. We've got technical computing, you know, large simulation clusters. You want to do checkpointing fast. We have ability to take these utilization of these clusters from, you know, 50% to 100% fairly quickly, not idling these 2,000, 4,000 node clusters for doing these checkpoints. So, multiple different ways are going to come. We're focusing on how to enable the channels, the end users, to really deliver the hardware. We've got a very unique software strategy on top. That'll wrap this up together. Well, as John said, it's a very hot area, right? I mean, it's always in any business, this room for more than one. And this is certainly the case in our business, right? And we've just fine... No, we've done a round of financing with Sequoia Capital last November. So, where are you at? How much have you raised? We've raised a little under $40 million. Okay, so you've got some... What's the round? How many rounds? We've done two rounds. And we'll probably need a little bit more. But we've got some very unique OEM deals I can't talk about today, but we'll see it in coming out in the next... Who else has invested? Just Sequoia or...? We have Sequoia, Globespan Capital, and Artem Adventures, 3 investors. Yeah, Artem and... They've been around. They're doing a lot of storage deals. So the early stage, they worked with us for the last three years. Where are they based? So they're all local. They're all Silicon Valley. Artemin? Yes. I never heard of them. Yes, they're... I thought I knew all the early stage VCs. No, Artemin is the early stage. Okay, well, thank you so much. Solid State's great. Entrepreneurship is hot. You guys are going to make a lot of money. Solid State's got a huge demand. I mean, just start pumping out the content, I mean product, because they need... They need... We need it. Shipping and they will come. We've got the best product today. That I can tell you. Okay, come on. Okay, we're going to bring Chris in.