 Live from Las Vegas, Nevada, it's theCUBE at IBM Interconnect 2015, brought to you by headline sponsor IBM. IBM Interconnect is a two hotel show at the MGM and here, people shuttling back and forth, keynotes are going on. Mike Kuhn is here, he's the Vice President of Flash Systems at IBM and he's joined by Tom Cook, CEO of Permabit. Gentlemen, welcome to theCUBE, it's great to see you both again. Great to be here. Great to be here, Dennis. I have a big couple of weeks for you guys. You had a big announcement last week in Alvedin, you're here at Interconnect, giant cloud show, 20,000 customers, so give us the update since we last talked. I mean, a lot of things going on. Great, love to, you know, two big announcements last week, right? We started the first part of last week with our announcements around software-defined storage, launched the IBM Spectrum family, a lot of traction around, you know, where the marketplace is moving, how we're responding to that, a billion-dollar investment around that part of the business. And of course, as you know, we launched a billion-dollar investment around Flash the lesson two years ago. So, last week, end of the week, we launched our third-generation Flash system, our all-Flash array, enterprise storage solution, put the whole software stack on top of it, complete integrated solution, highly scalable, it's just really been an amazing week for us last week. Yeah, so Tom, what do you make of that? I mean, it's like the, you know, we're from Boston, it's like the Brady to Edelman play that billion-dollar investment, right? You're making some investments yourself, maybe not quite a billion, but you've been at it now for a while with Permabit, really going after that data reduction space. Absolutely. Flash has been a big tailwind, talk about what's going on there. Certainly has been. We're probably the most focused organization in the world on data efficiency. More engineering resources going to it, bigger investment than anybody else. And because of that, we've been able to bring very high performance, very high-scale and very resource-efficient data deduplication compression to the marketplace. And we're happy to be partnering with IBM here on another offering, and that's what we wanted to talk about today. And you're right. Flash is, it's a big deal. The fastest growing products in the world today are all data efficiency enabled. And we think IBM has the best suite of attributes and the best suite of capabilities in the industry now. So before we get into the details of the relationship and the specifics, I want to step back a little bit and talk about some trends that you guys see. So our David Fleuer in 2009 did all these, he's a former IBMer, right? So he's good at, you know, he's a techie. So he's good at like log, log graphs and stuff like that. So he pulls out the data. 2009 he said, this is kind of an interesting trend. I think that Flash is going to actually be cheaper than high performance disk by 2014. So we wrote all this stuff up. People said we were nuts. And I don't think he missed it by that much. So we're kind of there, aren't we? I think we're there now. I think we're there right now. So he was right on. Yeah, I mean, maybe a couple of months off. Right. So, so we're seeing that high, that high, high performance disk is now an oxymoron. Right. So what does that mean? And what do we see for the rest of the market? So, so look, Flash is the, is the, is the fastest storage out there. Our clients are deploying about 20% of their data, their active primary data on Flash, the fastest tier of storage. And, and when we talk to them, we say, what's the other, you know, capacity layer? And the cheapest tier of storage they have is the storage they already own. And that's where our software defined strategy comes into play, right? So you can take our software defined capabilities, and you can, you can do better storage management, storage optimization on the storage they already own. And you can sort of seamlessly manage with our dynamic curing capacity, how you manage storage across, you know, both tiers of storage, right? But clearly, you know, we reached the tipping point with our, with data reduction technology today around the Flash business. It truly is cheaper than high performance spinning disk at 50x you know, times performance. So it's truly changing the way clients think about it. Of course, it's a very transformational technology, right? We lead with not the speeds and feeds of Flash and not the economics of Flash. We lead with how this is transforming their business. So whether we're talking about healthcare or finance or retail, we're talking about how it's truly changing their business. And now the economics have changed. It's truly been taken off. Yeah. So and Tom, when you guys first got into the whole data reduction sort of OEMing business, if I will, it really started with the spinning disk side. And, you know, you got kind of a tepid response, people don't want to put anything in between there. Now Flash completely changed everything and became a tailwind for you guys. What have you seen there? Well, certainly there were some early stage companies that set the trend here and got out there and built data efficiency into their Flash solutions. And when Mike originally acquired TMS, it was kind of just a screaming fast box, a tier zero box. And he's transformed into fully with full management features and data reduction compression technology of theirs. And what we've seen is the whole market now because the floodgates open, the whole market wants data reduction technologies across the board because it's all relative. Your savings are relative, whether it's spinning disk or whether it's Flash. And we've brought Flash together with partners like IBM into the cost spectrum that now customers can afford it for tier zero, tier one and tier two. And Jamie Thomas talked a lot about that the other day in her address, talking about really opening up more workloads for customers. This is just win, win, win for customers. And that's a great thing. So Mike, you guys got kind of a secret weapon in real time compression. You made the acquisition a number of years ago in store wise and it's worked really well. What are you guys doing together? What are you just piling on now? Well, as you know, as you know, we have we have two solutions. We have our flash system v9000 or a flash system 900. So the nine the v9000 is our highly optimized stack, single integrated solution, highly scalable environment. And then we have the 900, which is just a screaming fast flash array, right? And a lot of clients have said, geez, I'd love to have that tier zero application still, or I'd love to have it deployed in a way where I can just have, you know, data reduction on top of it, right? Due duplication, you know, real time compression. And that's where the partnership, you know, with permit came in, right? So it's a meet in the market play. It's me and the channel play. I was out here in Las Vegas about two weeks ago at IBM's part of world event. We talked with many of our distributors who are planning to announce this very, very soon in terms of having the solution available to our end user clients. Well, it's huge news because the knock on the tier zero was always great performance, but it's expensive. So now you're really attacking that. So what are the economics look like? How should we be thinking about that? Economics are phenomenal. I mean, when you think about with our real time compression technology in our V9000 today, we're getting about a five to one, you know, data reduction, right? So those economics say that we get to below $2 a gigabyte with our V9000 solution. Now with 900, we're seeing we've done some testing with the sand blocks product from Permabit and we're seeing, you know, six to one when you combine both deduplication in their compression and their appliance. And so when you start looking at those price points, you're now saying this is now truly cheaper than spinning disk, you know, not only at the acquisition level, then you start adding in all the benefits of, you know, better energy efficiency, you know, better cooling, less density. It's a no brainer right now. So last year, clients were saying, you know, why flash kicking the tires, lots of POCs. Now it's like, why wouldn't I do flash, right? It's interesting. I mean, I was on a road show last year talking to customers and I'd ask them how many people are using an all flash array and very few actually hands went up. That's changed dramatically. I presume you've seen the same thing. What are you seeing in terms of all flash array adoption in the marketplace? This is certainly we're seeing growth in the marketplace. The top two products in the market over the last year have data efficiency. They're both flash products as well. And Mike's product is right up there. And we certainly think deduplication will help him get that in the highest growth factor in the industry. It's what customers want today. And it's really a no brainer as a customer because the latency is so manageable and so superior to other storage environments that with the data efficiency, you have an extremely cost affected but very high performance solution. So you're a technology provider, essentially to IBM. You've got a pure sort of OEM model. What does it take to actually integrate the technology? Can you talk about that a little bit? So it's really as simple as hooking this up with your fiber channel switch and carving out a line and optimizing that line through us. And we work side by side the full stack of Mike's product, which is just great. Sandbox works with any fiber channel products. And so the promises will just work and it really does just work. But we've been working with the IBM lab for more than labs going back three years. And so these partnerships take a while to create and they take a lot of trust during that period of time. We've shipped nearly 10,000 units into the marketplace through other partners of ours. So this is a really a tested solution that IBM customers can really rely upon and and like in his organization. And Mike, I want to come back to sort of the IBM strategy. You guys made the acquisition of TMS a couple of years ago. You've married that with SVC or V9000. Right? Am I getting my product numbers right? IBM spectrum virtualized. Previously. SVC code stack. Yes. But that was that the core code that came out of SVC. And that was the stack. You've married those together. Correct. Excellent solution. But you had to have an SVC. Now you've got a full stack solution and your new announcements. And you've got a partnership with Micron that you're talking about. You're doing the world tour now. That attacks cost. It's modern platform, efficiency, performance. Talk about sort of that whole product line and that integration. So when we acquired the technology from TMS, it was a highly optimized hardware IO, hardware management platform, ran extremely fast. As you know, we're not using commodity SSDs. We're using our micro latency module. We actually launched it as the IBM Flash Core Technology. That's what we're calling it now. This is not commodity SSDs. This is our IBM Flash Core Technology, which runs at, you know, as you know, the microsecond speeds. So now we've married that with all of our SVC code stack. The reason why we don't call it SVC is we've done so much that. Sure, that's a 15 year old tested proven software solution. But that's been optimized for flash has been integrated in their single management interface. Highly scalable, scale out, scale up design. A lot of our value around our IBM spectrum family now is available single price point in our flashes and v9000 solution. So that really is our tier one play right now. And of course, we're still selling a lot in the tier zero space. Our flash system 900, which does not have the software stack on there as integrated, but you can add it on obviously. But that's also been a hot shell or as well, especially with our power systems brand. So power eight nonster cappy interface, right? And so where they talk about having two terabytes of memory and power eight, you can now because this is an FPGA design in our flashes 900, you can have like 57 terabytes of memory here. This thing runs so fast that it acts like memory at a price point that is order of magnitude cheaper than memory. That's a memory extension can be heavy and it's a sort of eliminates not only the spinning disk, but the overhead of a storage protocol, right? Right? So that would not be sent attached. That would be direct attached to a cappy interface card in the power eight server, right? And so you can deploy it that way. You can deploy obviously the 900 as a sand attached. You can deploy that with a solution like sand blocks from permabit. And then of course you have the V 9000, which really is the full integrated, you know, software stack with our, with our array. And that's sort of a game changer right now to get us into the tier one space. What are you seeing, Tom, in terms of if you hear this talk about the all flash data center? Are you are you seeing that either directly or indirectly through your OEM customers? We're seeing some in cloud. We're seeing some in selected spots, yes, or major sections of real time processing that's going on and things like that. Absolutely. But what we're seeing more is hybrids are the predominant portion of sales today. And that just means that a company like IBM can handle everything all the way through the stack right on to tape. You guys even reminded of that, that of us, us of that the other day. You know, the other thing I think is really cool here that we haven't really talked about is scalability. And in Mike's product really can scale, which is very different than some of the upstarts in the marketplace who've been out there. And you know, that's where the rubber really meets the road for the IBM type customer who's got hundreds of terabytes or petabytes of information that they want. And they want to manage it in a as as Eric Herzog referred to as an ocean of data rather than a lake of data. So that's pretty clever analogy. I love what do you guys think about that data lake data ocean? I haven't talked about data lake. It's a data ocean. It's a current in an ocean, right? It's more dynamic. We're riding the wave, right? You can ride the waves. I mean, I guess you can ride waves in lakes, but then, you know, not really. Hey, just to spend a minute and key on what Tom just said on the scalability of the product, you know, a lot of people talk about they have a scalable solution. They kind of have a, you know, small, medium, large solution, you know, with with with our IBM v9000 solution. It's this. It's a scalability design where you can start with one, you can scale to two, you can scale to three. It's it's seamless. It's non disruptive. I mean, these are the things that you would expect, right? Everybody talks about having a scalable solution, but only a few people really deliver what we just talked about. So that's a real differentiator in the marketplace right now. And I think that sort of is the value of IBM around this technology is you know, we've got everything from concurrent code load, concurrent maintenance, encryption, all the enterprise reliability characteristics, which you would imagine, as well as a very robust software stack, which has been optimized to run on flash. So we're really excited about that. But just let me say one more thing on that. So and this is one of the areas where most data efficiency solutions fall down. They can handle up to 20 terabytes or 50 terabytes or something like that. You start going into hundreds of terabytes or petabytes of information and they fall over. And it's because they're very challenging thing to do. And this is where the marriage is really a good thing because our capabilities really match up extremely well with the customer base and the capabilities of Mike's product line. Well, it's a lot of religious, you know, discussions going on out there and architectures, but you know, we're in the scale out big data world so we can see the writing on the wall. So I mean, everybody's got to get there. I got a couple other questions. You mentioned tape. Do you spend time thinking about tape? Tape's having a revival. Absolutely. You talk about the all flash data center. There are clients right now talking about tears of flash, hot flash, cold flash, no pun intended, right? They're talking about tears of flash. Tape's making a comeback, right? Basically, we see there's a there's a performance tier and there's a capacity tier and the capacity tier is making a big comeback with tape. So it's having a revival tapes keep part of our our IBM storage solutions. And we're starting to see that that has some some longevity with it. Reason I ask is you're heavily involved in flash and the marriage of flash and tape is something that we've looked at now for a couple of years, right called the floor called the plate, kind of a silly term. But the idea being if you have metadata today that's locked inside of the tape and you can escalate that up onto a flash layer, you can actually write algorithms to optimize the seek on tape and tape with flash becomes high both higher performance and of course much lower cost than disk. So is that a pipe dream or do you actually see clients thinking about making that sort of tape flash layer the sort of, you know, new bit pocket? Not a pipe dream. We're seeing a lot of this with clients that do security work and defense work and we can't talk about specific agencies. They're all three letter agencies, a lot of video surveillance and things of that sort, right? So, you know, having access to the data off of tape is hard, right? So using flash as the layer for your metadata to do some editing to do some real time stuff. And then of course, you know, demoting it back to the cheapest tier of storage is having a big impact from an economic standpoint as well as it's a lot more flexible than it ever has been before. Well, people don't think of tape as actually potentially faster, but discs, just spending discs, just keep getting slower and slower and slower. You can't spin them faster. Correct. Physics don't let you do that. And nobody's investing money in disk heads anymore. Why would you? I mean, maybe there's a couple of guys up there. A couple of six-year-old technology. Right, you can't move the data any faster off the disk. So it's like a little straw going into the disk where as tape, you can do a lot of things. Each hard disk drive is giving you about 200 IOPS per second. Iowa's per second, right? That's slow, right? Some of these flash arrays that we talked about, this 2U flash system, 900, 1.1 million IOPS. This is like an order magnitude over a spinning disk drive. So I want to ask you a question. We asked John and I asked this a lot of folks. We've watched a lot of transitions in the storage business. You saw the virtualization trend. You guys acquired XIV. Tom, you saw it. You saw a three-part, huge exit. You certainly saw it with Isilon. Great. Nice. So we say, all right, is flash going to be the same? Or if you look at the big players, IBM, HP, EMC, they've got their flash strategies sort of laid out. Will the upstarts, in your opinion, like to bolt you away in here, will they be able to achieve escape velocity? Or have the big guys got it right this time? You first. Well, you know, that's a hard one, because when we first started looking at all the flash startups, we were tracking probably 30, 40 viable solutions, right? When you talk about escape velocity. They escaped, but in the wrong direction. Yeah, I think you're talking maybe one, maybe two, right? That's kind of what we see right now. I think a lot of bets have already been placed. I think right now it's scaling very, very fast. You know, we talked about the stack, we talked about the performance. There's still a lot of people that are on commodity SSDs. They have to do something. We were doing commodity SSDs in 2009. We knew that that was only going to take us so far. So that's why we made a big shift two years ago, right? A lot of other people are going to make that shift again. Then you have to talk about service and support. I mean, we do the best service and support, I think anybody in the industry, and we do it in 120 different languages. So that's the kind of scale that you need to kind of take this technology mainstream. I don't think there's too many that are going to have that kind of escape velocity that you talk about. That's right. So the transformation that's going on certainly is kind of transitioning from where we are today with a lot of on-premise to much more in the cloud too. It's only the largest companies in the world that can really do that and do that extremely well. And we're seeing that really shake out. I kind of wonder whether we're going to see a real bit of a gap develop here between what the big people can do, the tier one vendors can do in the flash space. And I think they're going to take back the space that the upstarts have done over the last few years because they have the ability to build in their management stack. They're building in mature data efficiency, whereas the upstarts are kind of struggling to do both right now. So I think we're going to see Mike's company and others like him take back share over the next year. So we're getting the break sign. And my last question for each of you. Everybody said for years, we saw flash changes everything, flash changes everything. We certainly believe that. When we look at a couple years, when we look back, what's the big thing, Tom, that you're going to say, OK, flash really changed X? What is that X? So I think IBM has it right, business insight. It's going to give us the ability to really accelerate what data coming in and how quickly that turns into information that I can help operate my business with, as well as then tear it out and get it to a cheap space. So I think underneath it all, the facilitating technology, one of the facilitating technologies is data efficiency and it's mission critical to bring down the economics that we can do that. So you second that, Mike? It's really the business outcome? Absolutely. Let me just give you a little broader perspective. You know, speed matters, access matters, right? So our strategy around software defined around flash is really about having access to all the data. Data is exploding, as you know. And it's about doing things in real time. So it's going to be about driving data into the insights, right, and doing it in real time. This is what's changing people's business model, whether it's healthcare, whether it's financial services, whether it's retail. It's about doing things in real time that they couldn't do before with all these massive amounts of data. It's a data ocean. We're riding the wave. Haha, awesome. Well, storage remains the linchpin. Flash is really at the center of that real time. Mr. Gentleman, thanks very much for coming to theCUBE. It's always a pleasure seeing you. Thank you, David. Keep it right there, buddy. We'll be back with our next guest right after this. This is theCUBE. We're live from IBM Interconnect Mandalay Bay. Right back.