 Hi everybody, this is Dave Vellante. We're here at Wikibon headquarters and this is theCUBE's breaking analysis of the big news of the day, which of course didn't come as a surprise to anybody, but Lenovo, the number one PC maker, is buying IBM's x86 server business for $2.3 billion, $2 billion in cash and the rest in Lenovo stock. IBM is currently the number three x86 server supplier and behind Dell and HP or HP and Dell in that order. And with me here to discuss this at the Wikibon offices is Stuart Miniman, principal contributor and analyst at the Wikibon project and also David Floyer, the chief technology officer, David's out in Mountain View. Gentlemen, welcome and thanks for coming on. Thanks Dave. So as I said, this is a $2.3 billion acquisition of a company, so IBM's revenues in this area are 4.6 billion, so that's about a 50 cents on the dollar. If you looked at Dell before Dell went private, Dell was trading at a valuation of about 40 cents on the dollar. Lenovo was trading at about 40 cents on the dollar, so they're paying a premium relative to the big PC companies, but Stu, why did IBM make this move and is it a good move? Yeah, so Dave, as you said, this wasn't a surprise to anyone. If we've been watching the x86 business, overall it's been shrinking and IBM has really been hit on this. The two big growth areas for this environment have been really kind of a web scale companies where the ODMs, those Taiwanese companies have really been taking a big chunk of the market and the only exception to kind of the shrinking of the market has been Cisco. Cisco's really gone after what was traditionally IBM's sweet spot in the market, which was the performance area. I think back, IBM really created the blade server market. They were the leader in that space for a long time, but HP came on strong. Cisco then took a big chunk out of them and IBM just hasn't been able to recover. What's surprising is the breadth that this does affect. We were at the IBM Pure Systems launch and the Flex Systems, which is just the standard compute architecture, converged environment for virtualization and all x86 environments are all going to Lenovo. What IBM keeps is the power systems, the pure application, the pure data, and of course they're keeping their storage, but all of the x86 is all going to Lenovo. So IBM moving up the stack, is it a good move? Yeah, so I think it is a good move. I'm not sure how IBM competes. It really needs to be through software, which IBM continues to have, and cloud services and other solutions that they can build, and I think it gives IBM some more flexibility and IBM just didn't have low enough margins and have a differentiating solution. So it makes sense just like they did in the PC business once again to sell it off to the Chinese. Yeah, so David, IBM sold its PC business in 2005 for about 1.25 billion and has obviously been very successful with that. You're not as happy about this deal from IBM's perspective. Why not? No, I'm not. I was very much in favor of the selling off of the PC business. I thought that was a very smart move and in retrospect, has been a fantastic move with the margins and the volumes of the PC business have come down and Lenovo has increased its market share and increased its efficiency and it's a positive in the salivation market enormously. So from a Lenovo point of view, this is a good deal from IBM's point of view. In my opinion, this is a good short-term deal but a very big, strategic mistake in where system computing in general is going. Where it's going is towards completed, converged infrastructure with the hardware and the software and the storage and the networking all together in one box. And this makes them unable effectively to compete in this area. They can compete on software but they need other people to put together the hardware and the software. So they cannot make a complete stack except at the very high end. And that is the end which will be under most pressure over the next few years. But in my view, they had to take a strategic look at this and they needed to blend in the storage, make it a platform to enable them to sell their software more efficiently and more effectively into the marketplace and use it that way as a way of driving out costs, getting economies of scale, driving out maintenance costs on the software. And the way they're going now, they are going to be only able to provide their software as part of other people's stack and that will give them very little leverage indeed in the future. So let me follow up on that. So your basic premise is that the world is going to converged infrastructure. You've got to own that system in order to compete most effectively. IBM by giving up its x86 business only leaves its proprietary power series. And I presume you're arguing that IBM doesn't have the volume to compete and be cost effective in that area. Is that a correct interpretation of your premise? Correct. Okay, so now why can't IBM do what others do and just buy in the piece parts and put it together? I mean, well, I guess Oracle kind of does that, right? Oracle doesn't manufacture it, certainly it's own x86 systems. I mean, it sorta does. It incorporates x86, Cisco buys Intel chips. Does IBM really have to be in the low end server business to compete for converged infrastructure? Well, when you're looking at converged infrastructure on the server side, you need to, the chips are the common element. You buy those from Intel. What is different is how you put the recovery system together, how you blend that in with the storage. It's all of the microcode, all of the recovery system, those are unique to each individual server vendor. And there are a lot of chips that go into it other than the Intel chips themselves. So that's the piece that they will not be in control of. Somebody else will be in control of that total architect. It will put together the storage and the server and the networking as a complete package. And that means that they have no way of owning a whole stack. They can't buy in the piece parts because they won't own the microcode and everything else. They won't be able to put together that total stack. The people that will be able to do that are of the HP and Dell. And Lenovo, I would see that Livono will be going into the storage market to complete these deals. Well, Lenovo has a deal with EMC, of course. And don't forget Oracle. Oracle, obviously. Well, and actually as part of this deal, Lenovo will actually be selling IBM storage. So that does put into question how much the EMC Lenovo storage deal is going on because that was part and parcel of this x86 deal is there will be a strong partnership between IBM and Lenovo. IBM's going to continue to work on Windows and Linux development and they're going to use Lenovo. x86 is one of their piece parts. So I don't think they're pushing it off and saying we're not going to touch it. And now Lenovo is going to sell the storage. Yeah, so I mean this deal obviously was the head of the IBM server and software business is Steve Mills. He's a software guy. My view, he's never really loved hardware. He's not hated it, but he's a mainframe, et cetera. But he's a software guy and the x86 business has always been a low margin sore for IBM. So I think Mills looked at this and sort of disagreed with you, David. So strategically we got to get rid of this and focus on software and obviously services. Stu, what's your take on the conversion infrastructure piece? It's an area that you follow quite closely. What do you think about Flora's premise? Yeah, so it's that counterbalance. So I understand what David's saying and obviously HP's done a great job of bringing their server and their storage groups together. IBM hadn't done a great job with that. Dell, it's interesting if I look at them from a converge standpoint, they actually sell Dell servers into a lot of converge platforms. They're the primary partner for SimpliVity and there's other hyperconverged platforms that leverage Dell x86. But I really think that IBM can do more in software to be able to pull that together and that they don't need to own their own x86. They can partner closely with Lenovo and also have choice to use others. Interesting discussion we've been having out in the social spaces is if you look at IBM cloud, has there been too much of an incentive inside to use IBM servers for cloud which might slow them down and hurt their overall margins and if they get that business out that frees them up to do a lot more with it. So I think this is overall a good thing for IBM. They really don't, as you said, want to be a hardware company. It's their services and their software and they've got a robust software portfolio. IBM's server business has been in decline. It declined double digits last quarter and has been on the skids. The IBM CEO mentioned on the last conference call that they have a new mainframe coming out in 2015 and that should help shore up the hardware business. I found that interpretation of that comment is a little bit concerning. We're gonna wait till 2015 for a new mainframe to shore up our hardware business. David, does this mark the beginning of the end of IBM as a leader in systems and hardware? Totally, to me, the way that it's going, I mean, clearly they're making a loss on this part of their business and I understand the pressures on that clearly they give them a 20% over the odds on selling it. But this is a very short term. In my view, when you look at what Amazon is doing with the combination of software and hardware, they're investing hard into hardware and they're investing hard into the whole infrastructure working as a single unit and they're investing in taking out costs of the application running on that hardware all the way up the stack. That is the model that's going to work and IBM's service business, unless it adopts a similar model and it owns the whole of that stack, I don't believe we'll be able to complete effectively long-term with Amazon and other Amazon lookalikes which will come into that business. So it's very interesting, David, you're putting forth the notion of Amazon as a server supplier. Stu, thoughts on that? Yeah, so Dave, I wrote a big article talking about this point that David's making. Amazon is really focusing on hyper-specialization. They've got people working on power supplies and all the pieces. So the counterpoint I guess I would make is that IBM really looks at open source as one of their key strong points. They want to differentiate on execution. They want to use open source projects. Amazon leverages a lot of open source but doesn't contribute back to the community. There was a big article just written this week on the register talking about Amazon's actually losing some of their key engineering people because they don't contribute back to the open source community. So I think this is very diametrically opposed views of what Amazon's doing and what IBM's doing. IBM wants to leverage the open source and therefore getting rid of their own x86 helps them to do that. Well, what about that, David? Next week we're gonna be at the OCP summit. I'm hosting a fireside chat with some folks from Merck, OCP, and George Lessman at IO. You're gonna be there. What about that? Can IBM leverage its open source mojo and with initiatives like OCP actually replicate the capabilities of Amazon and focus more on cloud and jettison some of those financial losers? What do you think about that? Well, they could if they were able to put together that whole cloud strategy on their own hardware and open source software. I agree with you completely that they've got a great mojo with the open software. That's a great strength that they have. But the whole point that I'm getting at it is it has to be a vertically integrated stack to be cost effective to drive out costs. You have to have volume all the way through. And if you don't own it and you need other people to service it or you need other people to provide it, that won't work. So are you implying that initiatives like OCP because they're an ecosystem built approach won't be able to compete with fully integrated stacks? Not at all. The OCP is a fully integrated stack. And if you integrate that, for example, with open stack itself, you can create from that or people will be able to create from that a total stack. But they've got to own all the pieces. And if you try to get into it with bit parts you'll be a bit part player and your margin are going to be increasingly under pressure over the next decade. This is not going to roll out in a year or two years. This pressure on IBM and their move away from systems to a pure service player with a little bit of software they can sell to other people, I think in my view is the strategic mistake. Okay, we got to leave it right there. Thank you, David Floyer and Stu Miniman. This has been Breaking Analysis. I'm Dave Vellante with Wikibon. Check out siliconangle.com for all the news, Wikibon for all the research and Watch the Cube. We'll be at Palo Alto next week in Silicon Valley at the OCP Summit and also the open stack event down at the Mountain View Computer Museum. So we'll see you then. Thanks for watching.