 Live from Boston, Massachusetts, extracting the signal from the noise, it's theCUBE, covering Red Hat Summit, 2015. Brought to you by Red Hat. Now your hosts, Dave Vellante and Stu Miniman. Hi everybody, we're back at Red Hat Summit. Doug Baylog is here, he's a Vice General Manager, excuse me, of the Power Systems Group at IBM. Gave a keynote this morning. Doug, great to see you again, CUBE alum, awesome. Thanks for coming on. It's great to be on theCUBE again. Feels like just yesterday I was here as we were out at Edge out in Las Vegas. So not too long ago. You know, and I invoked Rosemilia's line on theCUBE the other day, infrastructure does matter. It does. Even if we want it to be invisible, it matters a lot. You guys have reorganized your group, really done something that I have been talking about since, you know, a long time, since Gershner's, maybe even before, is to get the right software products and marry them with the hardware, get the middleware in there, get the Tivoli pieces that matter in storage, bring it all together. It's a powerful story that you guys have. Set it up for us. Yeah, so, you know, we've been as a company transforming, why are we transforming? Same reason a lot of other companies are transforming. I mean, there's just tremendous forces that play in the marketplace at a faster and faster pace. We were just talking about a second ago. Between the shift to data, we're all swimming in data, trying to get any analytics inside out of the data, the move to the cloud. Of course, our point of view is most of our clients are going to be in some hybrid structure in the cloud. And of course, then the whole aspect of, kind of the way in which firms engage with their employees or engage with their clients, right? New workflows, new business models, new technology. All of this says, you know, we as a company, IBM have to keep transforming, and that's what we've been doing. To your point, David, the big shift we made in the next step of our transformation earlier this year was to say, you know what, let's bring together all of our capabilities around a key set of kind of growth areas, whether it be analytics, the cloud, Watson, security, commerce, now healthcare. Let's put those technologies, those services, those consultants together in the units. And then also though, let's take the infrastructure, the middleware, the Tivoli storage content, and bring all that together with servers for the first time that we've ever done it. So now we have a systems unit that I think really represents the platform of what we can do to bring the best of IBM software, systems, storage together, but still recognizing the role open plays as we go through all of that transformation. So a lot of change under foot, they're all sort of driven by market forces. Well, and you know, people like to take shots at IBM because it's big, IBM's getting out of the hardware business when you sold your X86 line, but when you sort of think about the organization, it's a much, much more powerful approach from a go-to-market, R&D services standpoint. New mainframe cycle, so Z is kicking in. Z13, huge first quarter, 130% growth in the first quarter. People don't understand how big this is. It's like kill, mainframe cycles aren't necessarily, you know, you don't see a new mainframe every year and the big mainframe customers, if you shouldn't be in the mainframe, you were off the mainframe a long time ago. If you're on the mainframe, you're going to buy that stuff sight unseen. It just throws off, you know, great business for both sides. You got a new spring in your step with open power. We're going to talk about that a lot. New storage strategy, so that's great. And the middleware piece, your Marie Week running that, you know, connecting in. So let's get to open power. You guys kind of doing a judo move on the big player in the industry using open source. I've never heard it called judo move. I'm going to use that line David. I mean, you know, here's IBM. Now IBM's got some, you know, you were very humble this morning. You said in our own little way, we helped, you know, establish Linux. If it weren't for, you know, Steve Mills vision, the ability to actually at the time invest a billion dollars in Linux, you know, the pace of Linux adoption and development. It would have been a different pace. So I think that's absolutely fair to say. Change is the game. Now, maybe it still would have happened, but IBM was an accelerant to that. Now you're trying to sort of use that same playbook in open power. Maybe talk about that a little bit. Well, actually use what we've learned in Linux on the mainframe over the last 15 years. I was there when we started. Tom and I were actually, Rose Millie and I were actually there when we made that decision to put Linux on the mainframe. Gosh, in hindsight, I wish we had done the same thing on power back then, but we didn't and that's history. But here we are today and obviously a lot has changed. And so it's not just about Linux on power. It's this point of time we see in the server business where we're going to really take all the concepts of the open source software and we're going to bring that to open source hardware. And that was sort of the aha moment we had with Google, with Melanox, with NVIDIA and Tiantz and you know what? Now's the time to disrupt it. And why? Because in the server space, certainly the distributed server space, it has been so much historically driven by a generation of chip after generation of chip after generation of chip. Now with the technology headwinds we all face, the, you know, the demise of Moore's law, however you want to think about it, you needed something different. Otherwise, listen, I haven't found an application that says, you know what, I'm okay running slower. You know, I haven't met that person yet. Maybe someday I will. So a disruptive play says, let's just open it up. Let's steal from, steal in sort of a cognitive way, right? From the open source software model and bring that to hardware. And that's really at the heart of the open power foundation. So it's Linux, it's little Indian on power. Basically bringing, welcoming in all the software. All the software, all the accelerators. I think that's the other big point, right? I mean accelerators. You know, if you think about what is going to, as a community, help innovation to go, it's going to be new models around chip innovation, new models around accelerators, high speed IO, IO adapters, clever system designs, flash is part of the memory space, new memory models, software optimizes. It's going to take a community. You couldn't do this. You cannot do this by yourself. No one company owns the innovation, should own the innovation agenda. Yes. And that's the breakaway model. So Doug, you know, one of the exciting pieces of the open powers, Google's involvement, can you give us a little insight and what brought Google involved? What did they see? Why did they want to participate? Because, you know, we've been tracking for years, you know, Google is, you know, kind of the typosphere is what we're going to see happening in the rest of the enterprise. Yeah, there's a bit of a common model without talking too specifically about Google because obviously, you know, I've got to be careful with that, right? But let's just talk about the principles of what brought Google in. Google, smart company, a lot of innovators, they saw themselves now sort of being single threaded to one chip provider, not their model. They want innovation, they want choice, they want to be able to move with speed, and they didn't feel they had that. So they were always looking for choice and therefore power technology, not necessarily an IBM power server, that's not their model, but power technology and in an open way was going to give them that ability to have choice in their infrastructure, to innovate at the speed they want to innovate and create that ongoing sort of competitive nature. That's what sort of brought them into open power. Now, the other thing I'd say is sort of the common thread you see through all of the cloud companies are standardization of their era of infrastructure is what matters to run a well-efficient cloud infrastructure. The interesting thing is all the clouds are standardized on something slightly different. That's why they go by from sort of the various white box providers, the system that exactly matches their need at the right cost point they want with the right logical interfaces. As IBM brings power to software, which is happening here in the next couple of weeks, that's not an IBM server. They're actually acquiring it through the supply chain that they would normally acquire it through. It's a power A processor, but it's an open power server. So, Google's buying from their supply chain, I suspect. IBM software is buying through their supply chain. RaxBase will buy an open compute based on power through their supply chain. It's about sort of creating the supply chain that serves the hyperscale data center space. Interesting. So is it almost like an ODM equivalent for power? I'm just trying to reconcile that. Yeah, it's a bit like an ODM, except the good news is because we've released the specs, we've released the firmware, we've released it all out there, these companies are surprising as at times. We had the Open Power Foundation Summit here in San Jose about a month or so ago. 15 companies came with systems. We didn't know that was going to happen. Same thing in China here. Just a couple of weeks ago, we weren't even at the event. It's 600 companies in China got together to talk about open power. And for the first time, they announced their own, what they called, red power. Local developed, local manufactured, locally built, locally innovated power based on Linux or Linux on power for the market in China. So we were starting to see this community-based, they just grab it and run it with it now. And that's very different than an ODM model. It says, I'm going to one-on-one work with you to bring a platform in the marketplace. Very different. I'm curious, does open power fit into the open compute discussion at all or something we might- Oh, absolutely. In fact, as I mentioned, so Rackspace who has chosen open compute as their form factor, as they're a member of the Open Power Foundation, they're now working with their supply chain to get a power-rate design system into the open compute model to bring in their infrastructure. So what I announced here, in fact, today, as you saw, we've got a number of advisors now to the Open Power Foundation. I shouldn't say we, the Open Power Foundation has a number of advisors. One of them is the Open Compute platform. The other one is the Linux Foundation and then the China Power Technology Alliance. They're going to advise us on what we should continue to do in open power to innovate and move forward fast. So a couple of other things I want to touch on from your keynote. You talked about, we like the Jeffrey Moore scenario, too, of systems of record, systems of engagement. You call the systems of insight, I think you call the systems of intelligence, but same concept. You also talked about, and just now, the attenuation of Moore's law. So it seems like innovation is going to come from combining technologies, combining these different systems, bringing transaction and analytic systems together and engaging with customers. So I wonder if you could talk about sort of IBM's vision of those three types of computing, bringing those together, powering those. What's the infrastructure look like underneath that? Yeah, I think what you'll find is they'll always be blending between them, but I think in many ways they can be very distinct. So the systems of record today, we're talking about the mainframe of the Z13 cycle. Fabulous technology, that is boy, what an example of systems of record continues to be at the largest financial firms, telcos, governments, insurance around the world, right? Same with our high-end power business, running AIX and IVMI. This is the final version of truth in a company, as I described. It's a really bad day when that stuff goes down, right? Because you're out of business across the board. So those records, they move at a very methodical pace. They got to be careful of the changes, change management's critical. Those aren't going to change. The systems of engagement are typically these days, sort of small-scale-out systems, and they're a rapid line of business decisions. Hey, let's experiment with a new application, a new workflow to go look at how do we, as we've done with Apple in our relationship there, the example is how do you allow a flight attendant, flights delayed, that happens sometimes I hear, right? Flights delayed, I'm going to rebook you while we're in flight, knowing that you're not going to make your connection. That's a whole workflow. That's not just a pretty app on the screen, that's a workflow back to the back end. So that engagement, and there's a experiment to get, try things, they work. I pay up for that. I told them don't do it, because I'm going to make that connection by the way. I said don't you dare change me. I've had too many examples of that. But those engagement applications, they're quicker to develop, they're experimental, you try stuff, you change it, often driven out of the CMO, the Chief Marketing Officer, move it a whole different pace, but often they're connected. And then the systems of insight, as we call it, kind of connects together that social world, the sentiment that exists out there in that mobile digital economy we find, with that data that resides in premises. I mean, grocery stores are a great example. They know everything about what we buy. What they don't know today is what do we say? Yeah. Right, so connecting together, what do we say in the social world with what do we buy? That gives them a whole different view of us as customers. Who wouldn't want to have that? So it's the combination of those technologies that really is driving innovation, not that faster processor speeds won't, but it used to be just that. Different design points, different architectures, data is sort of the thread through all three of those. So I want to also talk about it's not your space, but you talked about it in your keynote, Ron Perry, I know Ron, I've interviewed him in theCUBE a couple of times. We've had dinner together. Great story, isn't it? He talked about years ago the high-priced tyranny, the mainframe, how he swept his floor of all these x86 boxes, bringing in mainframe. He brought the tyranny back, I think. The most interesting part of his story, I can relate to it because I booked a flight the other day on some other airline, but it's serviced by, I won't say the airline because they're probably a customer and they'll be screaming at you. I could care less, but anyway, what his company is doing with data is essentially unpacking that whole value chain so that you can actually, as a customer, and I got five reasons why I couldn't make, when I called up, I tried to make a change. Oh, you can't make a change because it was booked on this company's paper, this airline's paper, and where this, and there, so he's unpacking that whole thing and allowing the airline industry to actually understand through data what's actually happening and what he's doing is bringing analytic and transaction data together. It's a very powerful story. It is. And if you think about the airlines he serves, a small and mid-size, what a great value prop for them as well too, right? He's not dealing with the large, multinationals that, you know, we often fly once you on their paper, right? I was, unfortunately. But I mean, he's dealing with sort of a bunch of carriers that want to connect themselves together, right? Their value versus the large ones is being in a connected economy. And so his value prop ties to their value prop of how do they come together as a community. So it's a whole different model. And that's digital disruption at its core. I mean, we all, you know, all those of us who fly some of the apps, they work great, right? You know, this is so easy. Other ones, it takes 20 minutes to log in. Yeah. You can even find out what it is. So that's going to change the way, you know, and that's just one example and, you know, airline transportation of bringing together analytics and transaction systems. So what are you seeing specifically, you know, one of the things we've talked about in the past is the tailwinds of big data on what used to be known as HPC. Right. And that sort of expertise into big data. What are you on analytics? What are you seeing there? Yeah. I mean, we as a company just made some big announcements last week in terms of kind of going all in on Spark. We do think Spark as sort of a disruptive technology is exactly right as sort of that next wave of the data, the language of data. And so, you know, the 3,500 developers we've committed to move to Spark. A lot of those developers are going to be on optimizing Spark on power. Right. So we've got the connection here beginning, so to speak, of that next generation of the language of data tied to the great value of profit power rate as the platform for analytics. So that's an interesting way back to your earlier point, David, of how we're connecting together some of these things. Big bet on Spark. You had a bet on Hadoop. You know, decent sized bet. Big, big bet on Spark. Well, I think back to the mainframe discussion. A Linux on the mainframe we talked about 15 years ago. I mean, we do have that position that when we sort of bet and go early in some small way, perhaps, right, we're able to influence the direction of the market. And certainly, again, we are betting big on Spark. And by the way, a lot of clients I've talked to are betting big on it too. They're excited about our contributions. Well, it makes the whole platform relevant. Yeah. You know, and let's face it. That platform was struggling with relevance and now all of a sudden, boom, you're throwing out names like Google, obviously Melanox and NVIDIA in their spaces and key, key players in the value chain. Now, 140 members of the Open Power Foundation. I mean, I just stopped quoting numbers because the moment I say it, it's obsolete. I remember when you announced that, you know, it was 25. It was 25. It was like, oh, there's some impressive names on there. But 140, you know, year on out is good. And they're not just Barney announcements. These are guys doing real development, correct? And I think, you know, if you think about sort of the waves of community to building. First off, you got to create a community. I think with 140,000s of ISVs on Linux on Power, tens of thousand open source packages, it's fair to say more to do, but we got a pretty good size community now. Next phase is demonstrating that innovation the community can do. And the two summits I refer to, one in San Jose, one in China, are starting to demonstrate that wild, crazy innovation that happens when you get people together. And then we sort of move to the monetization phase, which is how do we, we're all publicly traded companies for the most part. How do we then start to take that open innovation and bring it to market, bring it to clients as products and offerings and turn it into monetization. And I think we're starting to see that. You're talking your keynote about Watson again, another area that you don't run, but actually Watson, there's a couple of divisions that have Watson embedded. Pigeonos group, there's a Watson group, a whole group itself. You love it because it's an interesting use case, application, analytics package, whatever you want to call it. It's a great proof point for the value of power when it comes to big data analytics and a cognitive in this specific case. It's going to differentiate your platforms. But so you talked in your keynote about, yeah, many of you may remember Jeopardy. So Watson has really come along. It's got, like I said, two divisions within IBM embedded in one. Oh, even three now with the healthcare unit, which is grounded in Watson, right? So talk about Watson, it's like a secret weapon that IBM has, it's really unique. It came directly out of IBM organic R&D. IBM still invests a lot of dough in R&D, big differentiator for a lot of the whales. So talk about Watson and its importance to your group. Yeah, so certainly, as you said, David, if you look at the Watson team, they sort of worked on the grand challenge that came during our 100th anniversary as the Jeopardy machine, right? Was really a marriage between deep query analytics and natural language. How do you get an interactive sort of at that time, sort of interface that could understand human language, English at the time, right? And all the subtleties of an English language, but then turn that into based on the data source and the deep query analytics. They needed the best engine in the marketplace to run that on that was going to be able to return the answers, to beat the competitors before they hit the buzzer. They picked power because it was the best platform, not because we made them, it was the best platform. And they ran Linux on it as an open, fast, innovate set of operating environment. So that has continued now since the Jeopardy into the various businesses we've taken Watson, big focus on healthcare and medical. I talked about cancer research with MD Anderson today, but also in financial and retail and all those areas by which Watson has a cognitive engine and the build out of the portfolio around that brings sort of this unique data consumption. I mean, everybody's drowning the data. So to get a cognitive system that can help sort through the diverse inputs of data, find patterns that nobody even thought to ask about, that's real value to the users of Watson. So you said in your keynote, if you chose the IT business, because you thought it was going to be stable, you picked the wrong business. Right. The whole Chinese problem, but may you live in interesting times, we live in very interesting times, one where a co-operation is sort of the watchword. And so this is your 11th year sponsoring the Red Hat Summit. Yes. And then we hear, now we have Red Hat saying, okay, we've got this new old thing called OpenShift. We're going after things like WebSphere and the competing with BlueMix, yet your group very tight partnership with Red Hat. So talk about the partnership with Red Hat, what it means, we'll talk about co-operation a little bit. Yeah, yeah, yeah. So I think at the core of this and why we've been a sponsor for 11 years is, IBM and Red Hat together share a common view that OpenThis is the way in which you drive innovation in the marketplace. And so we stay grounded on those beliefs, those haven't wavered in the 11 years. And therefore, while we're going to individually choose where we partner or where we compete, and by the way, we're very comfortable with that given the various markets that IBM plays in, we're grounded in the same vision of what is the right approach to the market. And so that keeps us together. And it's been a great relationship. Red Hat recently brought their little Indian support to Linux on Power. They're building out Rev and the rest of the Red Hat stack on top of Linux on Power. So this is exciting. I mean, we all know the size of the participation of Red Hat in the market today. It's significant. And so my clients want Red Hat, I want to offer the full Red Hat stack. Follow up on that importance of Little Indie and running Linux on mainframe, running Linux on Power essentially means what to the customer that you can essentially run any off the shelf application that runs on X86, you can now run it on Power. One of the insights, and I'll give Google credit to this back to your question earlier, the role of Google here in coaching us has been, we had big Indian Linux, which meant there were technology differences, not right, not wrong, just different that made bringing application code and developing it on power different than Intel. And so Google's coaching was, listen, just get on with it and make it the same so that code that ran on Intel's higher level language code, Java and scripted language, you could literally just pick it up and move it over and it runs without change. That's what Little Indian allows. Now, if you're running a C++ or Fortran, there's a recompile, but as an example, we recompiled all Webster, roughly 10 million lines of code. We found two code bugs as we brought it from Little Indian on Intel to Little Indian on Power. Almost nothing. Do you have to shave their head? No shaving of the head required. No, you know, only half a head maybe, I don't know, but a whole different look. But my point is it made it the same, the same, easy to do, not something different that you're fretted about. And that's been the benefit of us moving to supporting Little Indian and Linux on Power. It's made that portability almost seamless. Well, with cloud, now containers, techniques like you just talked about, infrastructure matters, but it is becoming more invisible, which is a good thing, right? Invisible but differentiated. Back to your earlier point, infrastructure still matters, right? And that's what you want. You want it to be invisible, but where it can shine such in those analytics and big data kind of problems, you want it to fully shine through. And that's what we've been able to do. Yeah, I want my mobile apps to be super fast. Sure. No, I might not know what it runs on, but somebody somewhere down the chain does. That's exactly right. All right, Doug, I'll give you the last word. Key message you want to leave the Red Hat community with? Yeah, you know, sort of as I closed the Red Hat Summit keynote here today, my message was, again, to no surprise, first off, IBM's absolute commitment to open this. And why we fundamentally believe open, this is what will drive innovation going forward as it has in the past, even though we're faced with a tremendous set of changes and forces in the marketplace around data, cloud and engagement, open, this is going to transcend all of those. And we're all in, we really are all in. Excellent, Doug, thanks very much for stopping by theCUBE again. We'll see you at the next show, I'm sure. See you a couple weeks down the road. Appreciate the update. Yeah, good to see you guys. Take care. Keep right there, everybody. We'll be back with our next guest right after. This is theCUBE, we're live from Red Hat Summit in Boston. We'll be right back.