 Live from the Palace Hotel in San Francisco, it's the queue at the HGST Press and Industry Analyst Briefing brought to you by headline sponsor HGST. Here are your hosts, Stu Miniman and Jeff Frick. Hi, Jeff Frick here with the queue. We are live at the Sheridan Palace in downtown San Francisco at the HGST Press and Analyst Day. I'm joined today by my co-host, Stu Miniman. Hey, Jeff. Hey, thanks for having me and we're going to be doing an intro here talking about a lot of announcements that HGST have made talking about some of the big innovations in storage, talking about what's going on with hard drives, with flash, with software, and David Floyer is going to help us tease apart all of these announcements, but just to give our audience a little bit of a framework for what we're talking about here, of course the storage industry is going through massive amounts of change. Things like flash have been really allowing for new opportunities in new architectures over the last about six years, but it's not only about flash because we have greater capacity needs, so there were some announcements today that we're going to dig through talking about the capacity and then there's ones that talk about performance. But if I think back about 15 years ago, Jeff, we really saw storage go outside of the server. We created a network storage, basically sand storage attached networks and NAS, the file attached networks, and there was a creation of an entire stack of software to pull all these services together, create utilization, and over the last really few years we've seen things being pulled back into the server because we have a lot more power in our cores, in our compute standpoint, and flash is making us want to pull things closer to the application, which means that the storage needs to be near the compute. Earlier this year, David Flare and I put out what we called the server sand architecture, and what this does is a lot of the value that you saw from network storage being pulled into the compute layer, leveraging flash, and even more importantly, leveraging software to create those new services. One of the big announcements here from HGST is what they called a flash fabric, which really meshes nicely with our vision of where we see the data center going. So we're going to have a lot of good guests. We've got the president of HGST, we've got some analysts that are going to tease through the numbers, the various division heads to talk through all of these pieces and explain to our audience why this is a compelling message and important to where the future of storage is going. At this point I want to bring in David Flare, Wikibon's CTO and co-founder. We wrote the market forecast on server sand. David, thanks for joining us. Thanks very much indeed. A lot of announcements here that we have to kind of squint through. What kind of stood out for you most in this morning's news? Well, I think there were three major announcements that stood out to me. First of all, it was the flash fabric, as you said, and that's if you like server sand, their first entry into server sand, and that's a very exciting drive. The second is the 10 terabyte announcement of the Helio technology, the Helium-filled drives. And also what was very interesting about that was the issue that the whole of the marketplace has with the access density to hard drives. They're using the single magnetic recording technology, and that's very good indeed for reads, but the writes is a little problematic. So you're going to have to learn how to use this sort of technology in a new way and introduce that. And then the third interesting one was the active archive platform. Very, very dense application of technologies and enclosures, claiming to be five times more dense than others. Obviously, time will tell. But three very interesting technologies in their attempt to revisit the storage of the future. A new way of doing it, completely different from Seabate. So David, you've been quite bullish on flash technologies over the last few years. Can you lay out for us where are we in the transition from disk to flash? Obviously, one does not completely obviate the need for the other. So where are we with the market and where do we expect to see things go in the next couple of years? So if we look at flash. Flash now in a straight dollar per gigabyte is for active data, warm data, is about equivalent. If certainly if you add on compression and deduplication it becomes actually better than equivalent. So we're at very much a tipping point where active data is moving very rapidly towards flash. The benefits of flash are obviously performance and obviously bandwidth. But it is a technology that isn't in any way mechanical. So it takes all the advantages of growing very, very fast indeed. And you saw announcements today of 3.2 terabytes. Those are going to be much, much bigger. And the drive sizes of flash in the future are going to even be bigger than the mechanical ones. So from a price performance is coming down in prices, roughly equivalent. The interesting thing about flash though is it allows things to be virtual, allows the storage to be virtual for the first time because you've got enough IOPS and you've got enough bandwidth into it to extract the data in many different ways. So the most interesting thing about flash is that you can, for example, have a single copy of data and use it for OLTP and you can use it for your archive applications. You could also use it for your inquiry, big data type applications. All three based on the same data itself. So being shared of virtual use of that data for many different applications. That again is a mammoth reduction in the amount of data you require. So if you add on the ability to do deduplication and compression on it and add on this ability to virtualize that data and use it for both OLTP and for your data warehouse, you have some amazing reductions in the amount of data that you have to store. So it really is at a tipping point. And I think that HGST has been very, very sensible in doing two things. First of all, separating out flash from the drive, not going down the Seagate route. I think that's exactly right. And the other thing I think they've been very good is they're doubling down. They bought Viridun, they bought Velibit, they bought Stack. They're doubling down on their ability to enter the flash market in a big way. Dave, I wonder if you could talk a little bit about earlier in the presentation, Steve Bolligan, President and CEO, talked about this transformational time that we're in because of this explosive growth in the data. But really, instead of saying we've got to pick a road, they're going down multiple paths at the same time and really introduce this new concept of something in between kind of hot active data versus cold store data in this active archive. So I wonder if you could talk a little bit about kind of reflections on prior transformational periods and what you think is so different about this one and the path forward. Well, the amount of data is growing massively. It's clearly growing from all the sources they said. It's growing from mobile. It's going to grow from the Internet of Things or the industrial Internet from all the sensors that are coming out. And it's growing from all of those different areas. But what's interesting about their strategy is they're covering both bases. And there's clearly a demand for archive. And it's going to be on disk for as long as we can foresee because you want to write it once and you often hope to actually read it never. So it's a warm type device as opposed to a warm data. So that type of device, they're creating very, very high density, putting the six terabyte and then eventually 10 terabytes together with their own assembly and their own software around it. And trying to get to a much lower power capacity. What's interesting in racks is that power density is the thing that's actually limiting the amount of equipment that you can have in a rack. So that's their strategy there. Very, very promising strategy indeed. All right. So, David, one of the things I found interesting is, you know, where HGST sits in the stack. So obviously they have, they don't make their own NAND. They buy it from a couple of suppliers and they supply to server storage and cloud service providers are their customers. So some OEMs and the cloud service providers. You asked a question in the general audience about just strategically where they fit and how much value they can add. Are they the next Veritas? Are they the next EMC versus how they have to pair with their OEMs? So what do you think of things that they announced here that they called like device affinity and taking that viriden software and extending it? Where is the room for HGST in the marketplace? That, I think, is the question. I think the first thing they've done is differentiate themselves from Seagate and they want to add significantly more value. What happened in the Sand Marketplace is that EMC came in, they added 10 times more value per drive to the drives they bought from Seagate and they created a large marketplace of which they dominate. They dominate from the number of amount of revenue but even more they dominate from a profit point of view. So that was their strategy. HGST is trying to keep more of that revenue. They are investing in software around it in much bigger packages like the Active Archive. Very much like the Flash Fabric. The Flash Fabric is a straightforward replacement for Sand. So they are towing into that marketplace where they are going to be much more of a provider to a broader marketplace. Now they go to market strategies through partners and through OEMs, et cetera, and that's their stated market strategy. But long term, most of the revenue comes from added value of the stack up the stack and that's EMC proved that. So they are going to have to dance, in my opinion, a delicate dance between being an OEM supplier and trying to get into some of the higher level things. They want to be an OEM supplier at that level. Time will tell how successful that strategy is but it's certainly some toe dancing there. But to me, the Flash Fabric is the most interesting of all. The potential there in that marketplace to replace the sands with server sand is very exciting to me. So now that we've started down kind of the Flash adoption curve in hindsight now, do you see it tracking about what you expected? Is it going faster than you expected? How do you see it kind of mapping out now that you've got a little bit of data? Well, the server sand we forecast as being very, very gradual. I think we may have been actually too non-aggressive in our forecast. I think it's actually going to go faster. There is a very, very strong need. The biggest constraint on growth are the applications. Applications are the thing that uses data and you need those applications in place. So the applications they were quoting today that 75% of new applications in the cloud or big data type applications, those are the ones that are going to drive this sort of technology in the beginning. So it's clearly that the cloud is going to be a big driver of this. Enterprises coming up first of all to replace their current applications which don't need quite the same levels of performance as they did. But they will be driven because the potential of these new applications to drive productivity. There was a quote today of 5% and 6% being better. I believe that's a complete under-call. The potential for driving productivity by 30% is in my opinion what is out there if you combine the data warehouse and the application, and then drive them to replace unfortunately large numbers of people that sit in organizations making day-to-day decisions. Those are going to be replaced by data scientists coming up with algorithms that replace them. And that's going to mean a major, major change. That's the thing that Flash is really going to do. It's very exciting, a bit alarming, but it's going to come. Yeah, but at the same time they talked earlier today that the growth of the data far exceeds the growth of the people that are going to be around to manage it. So in fact, those guys might have more stuff to do. So David, thanks for joining us. I'm here with Stu. We've got a great lineup coming up with Jeff Jenna Cowitz from IDC. A whole host of execs from HGST. So we'll be right back with the next segment after this short break.