 Live from Las Vegas, it's theCUBE. Covering Edge 2016, brought to you by IBM. Now, here are your hosts, Dave Vellante and Stu Miniman. Welcome to Las Vegas, everybody. This is theCUBE, the worldwide leader in live tech coverage. And this is our special presentation, our fifth year at IBM Edge. IBM Edge started in Orlando as a pretty boutique storage show, you know, maybe a thousand, couple thousand people and it's grown into an overall systems, IBM Systems Business event, hosted by Tom Rosamilia, who's the Senior Vice President of the Systems Division, which now comprises mainframe Z power, open power and storage. Under Ginny Rometti's guidance and leadership, IBM has restructured its hardware business, as you know, sold off its X86 business, along with its BNT networking business. And what's left is the remnants of IBM's massive hardware business, which is a business that is in a sort of flat or managed decline with an agenda to actually grow. How is IBM planning on doing that? Well, Stu Miniman and I are going to be talking about that all week here actually today and tomorrow, wall-to-wall coverage of IBM Edge. Stu, two weeks in a row, it's always a pleasure working with you. We were at the Riverbed event last week. Our colleagues are out at Oracle Open World and it's good to be with you again. Great to be with you, Dave. Back here in Vegas yet again. It's been a little while since you've been here. I was here for VMworld a couple of weeks ago and yeah, this is a real interesting show. It's my fourth year doing the show and I love how seeing how IBM takes some of these assets or integrating it to the rest of the portfolio. So things like power and mainframe, well, it's a lot about how Linux fits into those, how they fit into the cloud, how data and analytics, things like Watson and cognitive computing tie into these underlying infrastructure pieces. So lots of dig in, some really cool users that they have in the keynote and a lot of them we're going to have on theCUBE, really excited to have the MIT Technology Review, Red Bull Racing, lots of great use cases that we'll dig into to talk about what's going on. So concomitant with this event, we've launched a new site, ibmgo.com is the digital experience for Edge and five other IBM events, including World of Watson, which is coming up in October, but check that out. So you'll see the IBM keynote channel, you see the CUBE channel, obviously the CUBE content is open and no login required, some of the IBM content is premium, but check that out. And Stu, let's talk a little bit about what's going on in the marketplace. So we know IBM has sort of pivoted, it's sort of de-levered, it's lower margin x86 business, kind of complete opposite of what HPE and Dell are doing. Dell and HPE are surviving in a lower margin business. IBM, under again the guidance of Ginny Rometti has decided that the margin is of course in software and services and cloud and it's really, it's shed it's lower margin x86 and some of its networking business. That's completely transformed the way in which people view IBM, partner with IBM and how IBM goes to market. So from a storage standpoint, it's all about flash, it's all about software defined and keeping the portfolio alive. From a standpoint of Z, it's about making Z open and from powers perspective, power started off with five partners, open power rather, and then IBM has really been aggressive under the guidance of Tom Rosamilia of driving power partnerships, I think 250 now, not the least of which is China. And you did a tweet the other day which I thought was very interesting, you said what's the bigger threat to legacy hardware vendors? AWS or China, 70% of the respondents said AWS which is kind of obvious, everybody knows AWS kicking bot and growing very, very quickly but China is interesting, why China's do? Why do you see China as a threat to the legacy hardware vendors? So Dave, one of the most interesting things, you and I talked about it from our studio a week or so ago, is look at the latest top 500. The biggest, the fastest supercomputer out there is made with Chinese chips. So China couldn't get the latest and greatest Intel chips, so they made their own chips. We understand what China's doing in networking and compute, expect to see China doing more in storage. So, of course they've got their own in-house kind of captured market that they always want to buy China but where are they going to expand their growth? The US is pretty much blocked, a lot of those China made products from getting into the US market but we think there's a real threat there. It's a threat to the US economy from those standpoints, some of the manufacturers and makers of innovative technology is how much will China copy what they're doing, how much will China just go beyond what they're doing? Well, open power is a fundamental lever there because China wants to be self-sufficient and will be presumably by the end of the decade it's going to have its own chips based on power and perhaps other technology. It's got its own version of Linux, there's a cloudera of China, there's a wiki bond of China, that's just sort of the way in which China operates. So that's very interesting, here's what's to me interesting from IBM's perspective. So one might think, wow, what's IBM doing? It's giving up its intellectual property to the world. IBM recognized that the expense, the R&D expense of competing with the likes of Intel and of course, I guess to a certain extent Oracle is very, very expensive. The way to do that and continue to innovate is to open up the platform, take the source code and make it available to the world. IBM's taking a play out of its Linux playbook in my view, it realizes that the money isn't necessarily in the chips and in the hardware but the money is on the software on top of that. IBM has made many, many billions and huge, huge profits in the Linux business and it was really the first to popularize Linux into the mainstream and it did that by saying we're going to invest a billion dollars in Linux and it successfully moderated Microsoft's monopoly. It's taking a similar approach. You heard Tom Rossimilia talk about how Moore's Law has hit, you really don't see faster clock speeds anymore. How are they achieving faster speeds? It's with layering on more cores and so people are doing sort of unnatural acts now to keep Moore's Law going. IBM figures, hey, let's open that up to the ecosystem, allow the ecosystem to drive innovation and we will lay our software on top of that, bringing those innovations into our hardware systems, our cognitive systems, our analytics systems, AI. So a lot of the innovations that are floating on top of the IBM hardware and IBM obviously from a hardware standpoint will have hooks into that. The other piece of it is IBM for years has had a global services organization which is a captive channel for its own hardware products. That's how it's been able to successfully fund innovation in its hardware, despite the fact that it's not necessarily a number one or number two player in all the markets. It's taking a similar strategy with cloud. Cloud is a huge channel for IBM, isn't it? Yeah, absolutely, Dave. And while this is an infrastructure show, we keep hearing IBM, how is it pushing up the stack? How is it getting into the application layer? In the open source, it's why they've got companies like Red Hat and Hortonworks up on stage there. Now, not like Dell and Hewlett Packard Enterprise don't also partner with Red Hat and Hortonworks, but IBM has had a long legacy of how they deal with open source. All of their different platforms are supporting Linux and that's a foundational layer for everything that they're doing in the infrastructure. And from a cloud standpoint, right. IBM makes good revenue in cloud. I wouldn't say that they are quite in the same category as an Amazon or a Microsoft or anything. Well, they're not. They're not, but they are a good layer there. When we were... A good layer, isn't it? A soft layer. Yeah, yeah. They've got soft layer, they've got blue mix. VMware made IBM the first partner for the vCloud Air network that they had. So, IBM understands the enterprise. They understand how to leverage open source. And when we say, how are you going to live in the future when infrastructure is being commoditized in general? IBM has a good story as to how they are going to create innovation and push new solutions in the marketplace. Well, let's talk about that a little bit. So here are Dell and HPE. Jennising, it's their higher margin software business. HPE just sold off its software business for $2.5 billion in cash. And HPE shareholder ownership in 15.1% of this new micro focus portfolio, which comprises HPE software division. So, let's get out of software, essentially. HPE will say we're not getting out of software, but essentially they're getting out of any software that doesn't run hardware. So, they're still in the open view business. Okay, great. Check. Dell has sold off, you know, under the guidance of Tom Joyce, a friend of theCUBE. It's many pieces of its software division. Yeah, the old Quest stuff. The old Quest stuff, and they're keeping boomy, right? So, okay, so they're keeping some of the higher margin business, but there's some security stuff gone or going. So, Dell embracing lower margin hardware, taking on, you know, willing to take on ODMs in China. Now, of course, with EMC and the distribution channel that EMC has, broadening its supply chain. HPE going after embracing some of those lower margin business. IBM not, IBM running away, not walking running from low margin business, much in the same way that Oracle ran away from the low end x86 business that Sun used to have, moving up the stack, realizing that software, cognitive, AI, cloud, machine learning, analytics, that's where IBM plans on making its margins. So, it doesn't have to, we've talked a lot about the need to compete as an infrastructure player with Amazon volumes, unless you have a unique advantage, which IBM does, obviously Oracle does, but EMC and HPE are arms dealers, you know, to the cloud, whereas IBM has a captive cloud channel and can get those higher margins, bringing its hardware onto the soft layer cloud. And we heard Matt Cadu, who's coming on theCUBE later on, talk about their stack, which essentially involved a lot of different pieces, including Aspera. We heard Tom Rosamilia talking about Ustream today. Ustream is what we've been using for years to power the CUBE platform, IBM acquired that. We heard him talk about Cleversafe, a move that IBM made to basically compete with S3. So, here's IBM bringing all these services into the cloud and focusing on a hybrid strategy for both on-prem and off-prem or public cloud. So, bringing the story to cloud, what's your take on IBM's overall cloud strategy? You've summarized it before, let's dig a little deeper into it. How does it stack up relative to the competition? Yeah, so Dave, I think you nailed it. The proper compare for IBM really is Oracle today. You know, as they build their stack, you know, IBM is, you know, I'd say more open than Oracle is. Doesn't mean that Oracle's not doing some open source out there, but, you know, Oracle builds the red stack. You know, here's our hardware. You know, it's built on Sun all the way up through the database. You know, IBM, you know, OpenPower, you know, Z, great interview I remember last year from this show, interviewed Walmart. And, you know, I saw Walmart at two big shows that I went to. One was Here at Edge last year and talked about, you know, Global Z Linux. So, you know, you think about a company that has kind of the global reach of Walmart and, you know, ZOS, you know, globally distributed as a major piece of how they built their infrastructure. The other show they were at was OpenStack. So, you know, obviously, you know, Walmart, you know, values open source need to build a lot of their infrastructures are not going to do it on AWS and IBM's a strong partner for them on that. So, you know, IBM, you know, builds all the stack with a lot of openness built in and stacks up well against what Oracle's doing. But, you know, both IBM and Oracle, they have their own clouds. Unlike, you said, HPE and Dell, which are going to be the suppliers to people like Microsoft for Azure and all of their service providers out there. So, you know, very different paths they've had, you know, very much focused on, you know, the application and moving up the stack for IBM. So, I want to talk about some organizational action going on at IBM. Ed Walsh was brought in a couple months ago to run the storage division. Now, the storage division at IBM has been a revolving door of executives, a variety, a number of people. Ambuj Goyal, who was, you know, sort of one of the architects of IBM software and analytics business, ran storage for a while. Jamie Thomas, who's now running strategy for Rosamelia, ran storage for a while. She's sort of a Tivoli, you know, a Tivoli sort of person that understands the software business. She came in for a while, helped architect IBMs move into software-defined storage. Ed Walsh is someone who used to work at IBM. We all know Ed Walsh as a startup maven, an individual who's really expert at taking struggling companies, shoring them up, repositioning them, re-strategizing, flipping them. He sold Avamar to EMC, he sold Virtual Ion to Oracle, he sold Storewise to IBM, did a stint at IBM, went off and got his MBA at MIT at the Sloan School that did catalogic, left catalogic, and IBM, you know, was able to lure him back. They made an offer, I guess, that he couldn't refuse. A lot of people have speculated, oh, they're bringing Ed in to sell the division. We'll ask him, even if I'm sure he won't comment on that. I'm not so sure. I think Ed Walsh, with his new, you know, MBA credentials, would like to be an IBM executive for a while, turn the division around. He's a storage guy, one of the few storage people that has been in this business, you know, run this business, Andy Monchah before, you know, the ones I mentioned, really wasn't a storage guy, even though he did the XIV deal, obviously learned a lot. But the people who ran the IBM Storage Division were more rising stars within IBM, where the storage GM was a stepping stone. Maybe it was like, hey, you know, we need an exec to shore this business up, you go do it for a while and then move on. So you saw a lot of that shuffling, Ed's a storage guy, so we're going to see him, I think, in that position for a while. The thing I wanted to bring you into is partnerships. Under Rosamilia, Tom Rosamilia, the senior vice president of the division, the IBM has really made a big push into partnerships. By jettisoning and selling its X86 business to Lenovo, that has opened up some huge opportunities for IBM. Now, first of all, can compete more aggressively with Intel. It also, as part of that deal, sold off its BNT networking division. So now it's got, you know, there's two edge sword there. One, it's got the no handcuffs now on partnerships. At the same time, the whole world is moving toward hyperconverged and they just sold off their networking division. So that means they have to partner for networking. And you've seen with VersaStack, IBM and Cisco getting a lot more friendly. You know Cisco really well. What's going on with Cisco generally and IBM and Cisco specifically? Well, Dave, yeah, not just the networking, I'll get to that in a second, but if I'm going to build a hyperconverged solution, what's IBM going to do? They're either going to have to partner with somebody like Cisco or they're going to run it on power. All the hyperconverged solutions today are built on kind of standard Intel processors, you know, Nutanix, VMware vSAN, you know, EMC scale IO, SimpliVity. All those guys are, you know, pull the storage into, you know, my Intel box. So it needs partners for that. IBM doesn't have an Intel box anymore and after they've jettisoned that and they're pushing the soft layer and they're pushing this, I wonder if they're just going to seed this part of the market because, you know, we've been talking to IBM for years and, you know, it seems to be that they would have a hyperconverged solution as part of their storage portfolio. They've done great in Flash, you know, they've got a number of pieces there. They understand software to find storage but maybe they just offer the software and they partner for the hardware, as you said, and right, if they manage the hardware piece from the compute, if they own the software piece either through an acquisition or something in-house, you know, networking is something that we're hearing more and more that you need to understand. If it's a very small configuration, it's not a challenge, but as you scale, networking can be, you know, it's something that breaks in the environment and something that they need to have. Something we've heard, you know, what used to be VCE, now the EMC Converged Platform Division feels that they have down-solidly, mostly through their partnership of Cisco. Nutanix, you know, I've talked to plenty of Nutanix customers that scale quite well and, you know, you just get people in there that understand networking to fix that. But, you know, right, IBM doesn't have hyperconverged yet and they don't have the networking piece and that means, you know, maybe they need to rely on Cisco more, which could be a little dangerous. Cisco's trying to go right into the hyperconverged market themselves. They've got their OEM solution called Hyperflex. It's an OEM that they have brought in through Spinkpath. And, you know, how are they going to fit in the mix? Cisco's done great in partnering with all the big guys, you know, IBM, Hitachi, EMC and NetUp unconverged, but for hyperconverged, all these guys are going through some churn and fighting each other a little bit as to, you know, who owns the hardware, who owns the software, who owns account control. Definitely something we've been watching closely for the last few years. Well, it's going to be interesting to see. So maybe IBM looks at, you know, networking, core networking is non-strategic. It might look at it as a lower margin business and or it doesn't want to take on Cisco's, you know, 66% market share. And maybe it's trying to leapfrog. So, some historical context here. Back when, years ago, when Bill Zeitler, this is over a decade ago, ran the IBM Systems and Technology Division, it was called at the time, he made a statement that the business is growing at 6% and R&D is growing at 9%, something's got to give. And one of the things that gave, and you've heard me talk about this before, is storage investment. So Bill Zeitler was a server guy and he said, I'm going to invest in the server business and I'm going to OEM a lot of the storage stuff. So it was LSI and it was MyLex and it was, you know, all kinds of third-party activity. It was bringing in IBM servers to be the IP of the compute portion of the storage. Remember, all storage runs microprocessors and so they really ratcheted down the R&D in storage. Well, you know what happens, when that happens, when that happens, your storage falls behind and then you got to play catch-up. Andy Monchah bought XIV because they saw an opportunity in virtualization, they had to fill a gap, they had to compete with the likes of 3PAR and Compalant and all those guys, so they brought an XIV, good little acquisition for them, made a couple of other tuck-in acquisitions, but IBM in storage, IBM invented storage. I mean it was the monster gorilla in storage for decades, but what happened was when EMC came in and knocked IBM off the top block, IBM had to play catch-up and IBM has consistently struggled in taking R&D and bringing it into product pipeline in storage and that's one of the things that we're going to ask Ed Walsh about, see if they can have been addressing that problem. Jamie Thomas addressed that through a software-defined storage strategy. What I'm getting too stew is maybe IBM says, you know what, this hyper-converged play is a chase that we're not going to be able to catch up because we're the tortoise, that's the hare, so maybe what we need to do is a leapfrog that with a true software-defined strategy through partnerships and with our own power strategy and go into more of a server-sand type of mode, like what Daytream's trying to do, like what E-Guaz is trying to do, maybe there's an acquisition for some of those guys. Remember, IBM is very, very good at acquisitions. It's just not been able to translate necessarily that acquisition prowess into the storage business. It's made some good acquisitions, but it's not hit the type of home runs that it has with, for instance, Acognos or its SaaS platforms that it's had over the years. So maybe IBM is thinking about a leapfrog strategy with regard to hyper-converged. What are your thoughts on that? Yeah, that's really interesting. So if you think back to the server side of things, you know, IBM Blade servers were the highest performance in the marketplace and when virtualization came, IBM was there, but Cisco, when they came out with the, it was the Project California, their UCS. Cisco passed them and really ate their lunch. And rather than try to fight back against that, IBM shed the X86 and focused on power. And if I remember right, the data that David Flawyer's put out on power, what Oracle's doing with the latest generation of their chips has higher performance than anything on the X86 side. So not following down the same kind of leapfrog that everybody else is doing, but using other resources and, as you said, trying to leapfrog what's happening there. So it's interesting, absolutely. Software is the real driver in storage these days. IBM understands what they're doing with cloud. They've got open source, they've got lots of partnerships. So there's more than one way to skin this cat of the storage transformation. Yeah, and IBM, with its pure systems, made an attempt at converged infrastructure. I think generally speaking, you would agree, it was not a home run. I mean, they made a run for it. They had great, great messaging. Steve Mills, I was there when they announced it, and I said, wow, these guys really nailed it. They're attacking the IT labor problem, but they weren't ever able to execute on the product and the go-to market. And then when they sold off the X86 division, that sort of threw everything up in the air. So I'm concerned about over rotation into server sand. I mean, we love server sand. We've put out these long-term forecasts. By the end of the decade, you're going to start to see the sand market really start to get eaten into. That's at least what we're saying. We'll see whether we're right or whether the IDC forecasts are up and to the right are correct. We'll see. But I think I'm concerned about companies over rotating. That's why I like the startups that are doing this. We mentioned a couple before, and maybe IBM's just sort of sitting back and saying, let's partner to fill the gap and let's watch what's happening. The big issue here is can IBM continue to drive innovation and drive growth to support its overall company strategy of cognitive analytics and so forth. And that really is the fundamental strategy of what IBM's trying to do, is support the digital transformations that are going on with infrastructure for the future. It's got a lot of good use cases. We've got some folks from the insurance business. We got some, you know, the CIO from Red Bull Racing coming on. IBM Edge always has all IBM shows, always have great customers, really good use cases, excellent executives, a strong women in tech presence. Of course, it's, you know, CEO of one of the largest tech companies on the planet is a woman, so that's, you know, high marks for IBM on that. Always really, really good content. I'll give you last thoughts on the opening segment and what do you think? Yeah, so just close out on the server sand piece. One thing to look at, it's not just about the directory placement for kind of storage that we have today. It is some of the transformational things. So if I can build infrastructure that SaaS can live on, that might not be in my data center, my customer might not be buying it. So that's where things like Blue Mix and Watson and all the analytics plays that IBM has. So they might look at the overall market, where it's growing, where they're going to be fighting for margin and deciding which ones they want to play and which ones they don't. That's like the clever safe acquisition is there's a storage play, but there's very much a cloud play. So, you know, IBM looks at, there's the whole 3D chess going out there in the marketplace, lots going on there. And as we said, you know, infrastructure, you know, is interesting, but it's the foundational layer to enable everything above the stack. Well, you know what, let's stay on this for a minute. I think we got time, right? We got some time. So, okay, let's stay on this, because under Gerstner, IBM made the choice of getting out of some lower margin business like spinning disk. They sold that division off to Hitachi. They sold, you know, later on, they sold the microelectronics division. They actually paid somebody to take it. It's like the Red Sox getting rid of a bad player. And, but it really shores up the balance sheet. It cleans up, you know, the income statement. And so, my point is that IBM is using a lot of its old plays that worked well, Blue Mix, you can relate to the web sphere success. Cloud is today's version of services. It's just cloud, it's services at scale, at repeatability, but it's a captive business. The difference is bringing in a lot more partnerships through technology, although IBM Global Services, I'm sure, installed many, many EMC disk drive over the years. So they would do that if the customer wanted, if the customer didn't care, then they would just integrate their own products. So you're seeing that play. IBM's billion dollar bets, the Steve Mills billion dollar bet in Linux. You've seen that in a number of areas, software defined, Flash, I'm not sure Flash was ever announced at a billion, actually Flash was announced at a billion. Spark, they never announced a billion, but I'm sure it was many hundreds of millions that they're investing in Spark. So you're seeing that IBM place its bets in a big way as it has done in the past, and it's worked really, really well for the company. So you're seeing this very large company slowly turn and now aim for the future. And I think you're going to start seeing, I've been saying this now for years, I've been sort of waiting for this to happen, but IBM's start really clicking on all cylinders, certainly from a financial standpoint. And it's been buying time with things like stop buybacks and dividends and so forth. And that's worked for the company. Now it's all about onto the future, which is cloud, certainly analytics and big data. IBM really doesn't use the big data term, but certainly analytics, cognitive IBM is way ahead of most enterprise players in cognitive, maybe not ahead of Google and Amazon and Apple, but they're with them and applying it with their deep, deep industry expertise. So, the future I think is bright, but still a lot of questions remains too. Yeah, absolutely. One of your favorite things to talk about Dave is the marginal economics here. Is IBM still too big to be successful in some of these challenges going forward? There needs to be that shift from services, which is very much throwing people at it, to software where you can gain that scale. And that's where you mentioned some of the big software players. I mean, even people like Facebook, Salesforce and the like are really starting to get into that kind of cognitive and artificial intelligence space. And can IBM's Watson be a big enough player there to be able to grow there? And how much, how many people do they need for that compared to traditional IBM global services, which has been such a big piece of IBM's business for so long? Well, SoftLayer is the platform for that. And BlueMix is the secret sauce to bring together IBM's disparate software business, because it is a very fragmented stovepipe software business that IBM brought in through many, many acquisitions, dozens and dozens of acquisitions over time. But so BlueMix is that secret sauce to bring all those things together, but you're really starting to see the IBM strategy coming to focus. Cloud, cognitive, scale, jettison low margin businesses leveraging its deep industry expertise. So, Ginny remember ran strategy before she became CEO, so she's got a pretty good handle on all this stuff. All right, Stu, that's a wrap of our intro. We are going wall-to-wall coverage of IBM Edge. Of course, our colleagues are across the way and in Moscone at Oracle Open World, John Furrier, and Peter Burris are there. We'll be covering Edge Live Monday and Tuesday, wall-to-wall coverage. Keep it right there, everybody. Check out ibmgo.com for all the IBM main tense sessions, general sessions and keynotes. And also you'll see the Cube channel there as well. And tons and tons and tons of great contact, great content coming flowing in from Silicon Angle, flowing in from Wikibon, flowing in from third parties as well. Keep it right there, everybody. This is theCUBE. We'll be right back.