 Live from Las Vegas, it's theCUBE. Covering InterConnect 2017, brought to you by IBM. Okay, welcome back everyone. We are here live in Las Vegas for IBM InterConnect 2017. This is theCUBE's coverage of IBM InterConnect. I'm John Furrier, my co-host Dave Vellante. We have Kayleen Sanchez, Vice President of IBM Enterprise Storage Development at IBM. We had an interview at VMworld last year about tape, making tape cool. Great to see you again. Thank you, and thank you for welcoming me back. So I guess I was in too bad last time. No, you're good. We love tapes. There's a tape culture out there. There's a tape community. Yes. Tape has been dead forever. It's going to die this year, as everyone predicted. They're going to die next year. It never dies. Tape is always around, and Dave and I, we see this all the time. Yeah, it's back. It's cool. It's relevant. It's the least expensive storage. That is correct. Out there. So what's the update? What's cool about tape this year? So I think when I was speaking to you earlier, you talked about FLAPE, what we're doing with flash and actual tape. So in partnership with our microcoders, our engineers and scientists, we partner in Tucson, Arizona with a team in Zurich in research to really figure out what we're doing with FLAPE. And by the way, FLAPE is a cool name, right? It's a very developer name. Well, you know, Wikibon coined that term. That was David Floyer. It's a flash plus tape, yes. But the premise was that there's not a lot of innovation going on in disk drive heads. Correct. And they're hermetically sealed, whereas in tape, you could do a lot more, more bandwidth, and you could do some cool stuff with search, right? And new tape formats, right? So that's all coming together. And are you, is there software now associated with that index so you can more quickly search? We have created a management layer that supports what we intend with FLAPE. And also across the tape portfolio to really consume applications at a higher level to enable what we need to do with our consumability, not only from a tape perspective, but also with flash. Right, so the economics really still favor tape, FLAPE, the flash just supports the speed. So it starts to encroach on some of that long-term archiving. Which is important based on archive, because aren't we all data hoarders? We like to keep our data and archive it and stage it off, whatever it is. It could be based on what we're doing with tape, and also hard disk drives. Some clients that I work with substantiate like archive data to like cheap drives as well, right? So hopefully eventually the transition will be to enable what we want to do with flash. Once flash of course is a cheaper or a stronger price competitive thing. So bridging though to from our last conversation, believe me, tape is sexy. So I'm telling the audience here, it's like, if you're not talking about tape, well where have you been? But at the same time, I want to talk about flash and what we do with DS8000. So we have an enterprise monolithic system that covers like six nines of availability that substantiates what we do in the flash market. And we just recently announced from enterprise down to entry, so mid-range as well as entry devices that are all flash. So we care more from an IBM perspective on what we're doing associated with the flash investment. Friends of mine like Eric Herzog, I sure get on stage with you to talk about like, IBM is focusing on flash. It's relevant to us. It's relevant to our flash. And the software too. Very software driven, flash is key. But you're bringing this capacity at the six, what was it, six nines? Yep. I mean. It's reliability. I mean that's just like just dump all the data. Perfect scenario. Yes. It's a beautiful thing. In addition to what flash does from an engineering perspective, forgive me I'm going to be geek for a moment, is that it allows us in the lab to focus on other things. So basically that latency or the chase for performance equals more. More meaning that we can focus more on what it means to develop optimizers, like for instance, easy tier, et cetera, to really enable a better benefit. And also some of those engineers and scientists allow us to focus more in Flap as well. So explain that concept, okay, latency equals more. What specific do you mean? Like the latency on the devices, the data movement, what do you just double down on that for a second? So from a performance perspective, we have to work around bottlenecks. That was where our focus was. But now with flash, we worry less about those individual components from a reliability perspective, as well as chasing latency or performance measurements based on IOPS. And in order to do that, we don't have to worry about it anymore from an engineering standpoint. So it allows us the time to really focus on what matters next. Like the value that we could think of is that we could benefit clients with regards to advanced technologies, technologies of value. Like what does that for you? It's liberating, it's creating community. Big data and analytics. Really, okay. Yes, so like for instance, the Expo floor associated with Interconnect, you meet all these people that talk about how data matters. Well, it's the intelligence around data. And so we want to figure out how to harness that data and drive out the intelligence or the smarts associated with the data. And that's what it allows us to talk about. Yeah, so let's keep on that theme of business impact. You were talking about latency before. Everybody knows flash fast. You're implying or I'm inferring from your statements that it's still more expensive. Correct. However, you got data reduction technologies and you have this data sharing notion. In other words, I can share much more data with the same copy out of a flash. That has an impact on developer productivity which ripples through on innovation. Correct. Are you seeing that have business impacts within your client base? Yes, so for instance, at first, we started to talk about, from an engineering perspective, compression and de-duplication had to be much more efficient with that data storage. And then afterwards, we started to talk about, well, we had to move quickly to serve our clients with those feature functions. And now we're about talking about how we harness or archive how we enable big data. And also the IoT aspect of the intelligence of the data and how we can translate that to improving client value. For example, I just saw on the expo floor partnered with Glassbeam and with the, I did basically a meeting with Glassbeam on the floor to talk about what we've done with them in partnership to harness the power of data. What is Glassbeam? So Glassbeam basically makes sense of the mess of data that we just have out there and makes it much more intelligent. So it allows their system, their algorithm to ingest data and better understand where we're at with that data. It's a new different than what we do with Watson as well. So with that, from a Watson perspective, you ingest the data and you can provide additional smartness about that data as well as the intelligence of. And what kind of data are we talking about here? Structured data, unstructured data? Structured data. And specifically associated with Glassbeam, it was all about really bringing in the plumbing of data for our call home clients worldwide. So clients experience our systems worldwide associated with DSA 1000 and we wanted to better be in a position to serve our clients adequately. And what I mean by that is they could have an error that occurred or we wanna be proactive with them based on call home and also some of the heartbeat information we get based on the systems and we wanna adequately share with them that. So you as a client could, I could send you a really beautiful simple email or communication, maybe it's a tweet that basically says, hey, there's something more worried about and we've got to proactively address it ASAP. Well, and there's all kinds of metrics buried in those files, right? There's utilization data, there's data on the effectiveness of thin provisioning. You mentioned compression, deduplication. Yes. I mean, I don't know what else is in there. There's probably a ton more stuff, obviously problems that occur. So have you been able to get to the point where you can be anticipatory and head off, front run some of those problems? So our end goal is to build an autonomic system. An autonomic system that has the brains to self heal. And that's what we want to focus on in the future. Now, are we there yet? No, we're not. But what we're doing with Watson or Glass-Bean or some of these optimizers, these tools to build better systems is something that we're doing associated with building the future of an autonomic system. I mean, one of the things John and I have been talking about with this, Ginny was talking about cognitive to the core this morning and this cognitive world we live in. Just a whole new set of metrics emerging in KPIs. I mean, you mentioned self healing, right? We still, to this day, track availability and okay, the light on the server versus the application, things like that. We see, and I wonder if you could comment on this, a whole new set of KPIs emerging from the infrastructure standpoint of what percent of the problems were self healed? And how can we affect that and increase that and what are we doing with that free time? Are you hearing that from clients that they're changing or adding to the metrics KPIs? Yes. So first, am I hearing from clients on that? Yes. So it's always these questions of like, okay, so from a cognition perspective, cognitive focus, what are you going to do to help us? To self heal as well as how do you build in the intelligence based on artificial intelligence to really self heal? And that's one of the focuses we're working on. What's the coolest thing happening now? Because last time I loved the consciousness we had about capacity and stuff that I learned was all the engineering, just to get squeezed more out of it, because the tape is a great thing. The reliability is killer. You got some great reliability, so it's a good solution. But there's always the engineering side of it that's science. What's going on that you guys are kind of digging away at and pounding away at for tech that people might not know about for tape? So using the cognitive systems or AI as the foundation, we're thinking about how to build in intelligence within our systems. And the way to do that is the reason why I keep focusing on this word, autonomic. How do we build a true autonomic system? It's almost like a system that has its own brain, right? And that chip set that exists inside associated with DS8000, it's like power devices, right? Whether it's six core, eight core, whatever. How big of a brain do you want? It's kind of discussion to have. But what's important about that is we really want to figure out how to be smart enough to self-heal. And we don't know how to do that just yet. And it's going to be just like you had mentioned, all this information and pulling it in to really determine how we go about doing so. So that's kind of near term. Those are sort of maybe in the binoculars, you can start to see how you can utilize analytics and cognitive to do some of that self-healing. I wanted to ask you a sort of telescope question. We heard Ginny talk today about quantum. What are your thoughts on that in terms of the implications for storage? My thoughts on quantum. So first of all, let's figure out how to harness the science of quantum computing, right? So that's the first like fundamental like, I don't know, first step of the 12 step program, realizing that challenge, right? So from that, it's like really realize that and recognize that and IBM is working on what we're doing with quantum computing. As far as how it relates specifically to storage, so we think it could be a benefit with relates to DS8000 tape as well. Because think about it, tape as far as the library side, that's what we did is we built out like infrastructure that really harness this aspect of data and did it in the cheapest way possible and in an energy efficient way possible. So I think quantum from our perspective is like a leap frog into the future of what we enable with some of our thinking there. And Ginny and team as well as her senior leadership are influencing how we should think about quantum computing as it relates to storage. So I say the next time that we meet, you should probably ask that question of me again. Like how far away are you? You'd be at step one and a half or two of the 12 step program? I would say one and a half. Go ahead, sorry. Go ahead. I want to ask you about what Ed Walsh took over. By the way, I like the two of you competing on questions. Yeah, we both like to talk. We can't get enough tape. We got tape everywhere, look at it. Tape me down the lights. So, but here's my question, Kevin. So when Ed Walsh took over the GM of the storage business, I asked him this. IBM's always had a rich heritage of R&D and development. However, my comment was sometimes it was sort of development for development's sake. And I feel like, and he sort of said this, one of my missions is to get, you know, aligned engineering with, you know, go to market, get stuff out of the pipeline into the market sooner. From an engineering perspective, have you guys begun to do that? What, you know, changes have you affected? Are you seeing the effects of that sort of initiative? So we have an agility process within IBM development that was basically, Ed Walsh was a huge advocate for that, supported it. And his intent is for us to push all of this wonderful IP that we build in-house to the marketplace as quickly as possible. So I'd say at this moment, we're there. I just, right now, he's in the nicest way possible in the most charming way. Telling me it's like, you're not fast enough, right? And that's a good thing. Like that means that there's more innovation, more intellectual property we can put into the marketplace faster, quicker, whatever that means in larger increments versus it being me. Previously, I told, I would tell you, it's like, ah, so DS8000, I may deliver that to you target-wise. It's 12 months from now. It's not good enough anymore. So Ed's coming on tomorrow, so we'll ask him how Caitlyn's doing, maybe. We'll put him on the spot and you on the spot at the same time, if you don't mind. Oh yeah, no problem. Caitlyn, always great to chat with you. Love these conversations. Thanks for coming on theCUBE, sharing the insights on the Tade to DS8000. Appreciate it. Thank you very much. This is theCUBE live here in Las Vegas for IBM Interconnect. I'm John Furrier. Dave Vellante, you're watching theCUBE. Stay with us. We've got more great interviews for the rest of the day and all day tomorrow. We'll be right back.