IBM's sale of x86 servers and networking to Lenovo

I’d waited a while before posting on this for a number of reasons, not the least of which was I was quite busy. But also, I wanted to understand what was and was not sold. Now that some of the dust has settled, and both companies have publicly discussed this, we know pretty well what is included in the sale. I don’t need to get in to that aspect, you can read it all very succinctly on Lenovo’s site. My first reaction to hearing this news was questioning the impact upon customers of IBM System x systems. The second question was, what will IBM do with all its Linux investment … is it pushing development off over to Lenovo? The Lenovo texts suggests that x86 development will continue, but I am guessing that this is in a transition to a services-like offering (WAG on my part) for development. Understand that I had thought that there might be a non-zero chance of IBM purchasing Red Hat in the near future. So seeing this change suggests that this concept is less likely. Unless that is, IBM wants to own the OS and platform and not the hardware. IBM has also signed Lenovo on to sell Storwize, GPFS, and a few other things. Things that make sense on x86, but maybe not so much on Power. This also frees up IBM from being a large Intel customer. Which might enable them to look at competing with Intel. Though I am not sure how … Power? PowerPC? I am not convinced that IBM wants to remain in the low margin hardware side of things. So I consider this path unlikely. A few obvious things about this purchase, and one, what I would consider massive, unintended consequence.

First, the obvious things. Lenovo will now have a presence on the top500 and other “metrics”. No brainer, it will happen. Lenovo will now be in the mix for full stack HPC and other clusters that used to be sold by IBM. More on this in a moment. Lenovo will now directly compete with Dell, HP, and the few remaining server vendors. Including Quanta. The Quanta/Lenovo competition will prove to be interesting. I used to say you can’t out-Dell Dell. This was based upon experience watching them in competition with others, as they sucked the oxygen out of the room. Well, I think Dell has just been out-Dell’ed. I’d also say that the vanity free machine market could get much more interesting, and that Supermicro has got to be on someone’s acquisition radar. I am betting a big mainland Chinese firm of one sort or the other. Those were the obvious things. I find it terribly interesting that, given how hard Dell, HP, and IBM have been wanting to exit this market, that IBM was actually the first to be able to do so. With the extraordinarily aggressive onslaught I expect from them, its hard to see how Dell and HP have a long term future in this market. I am not sure they have enough to sell off. Is it possible that the might simply shut those units down? In pieces if at all. I wonder what, if anything, now will happen with the ULLtraDIMM based system x units. Those are highly specialized units. We think they are targeting our siFlash, though we expect to be able to offer this functionality there as well in due time. Lenovo tends to focus more on volume plays than niche plays, so I would not expect to see much more investment in that route. Here is the massive, un-intended side effect of the sale of IBM’s x86 division and product lines to Lenovo. They will play it down, but, really, the magnitude of this impact is, I suspect, huge. This division would no longer be in contention for any politically or security sensitive acquisition. So large scale systems procured by US Government agencies with a security requirement of some sort, will likely need to pass over the IBM/Lenovo choice, by law. More to the point, IBM is terribly good at lobbying the NY State government to build projects. This is good for the state and for IBM in some ways. x86 based projects are now, very hard politically for NY state to accept. As Lenovo is not a US company. How would that look to the politicians? How would their electoral challengers portray them if they let purchases happen for NY state projects occur, from a “foreign” company, even if its, you know, “IBM” (scare quotes intended) though really Lenovo. This would be true even if the purchases happened tomorrow, or any time before the full closing. And its not just NY state. Its many places in the US at the state and federal level. If its x86 based, you have an optics problem. Most state and federal procurement require a “US first” purchase. Buy from a US company, parts made in the US. They have had to relax bits of this over the years, as drives and chips aren’t made here much anymore. Nor boards. But it would be hard to relax it for the entire product. Even if the systems were assembled here, the company ownership question is strong enough to force favoritism for US manufacturers. I am sure some will disagree, and I think we need to see how this all plays out. More interesting is that, now, IBM is competing against EMC and Netapp with storage, and big data appliances, but without servers. That albatross has been moved over to Lenovo, whom will probably turn that into a pot of gold for themselves. EMC doesn’t have big data appliances, nor does NetApp, or Cisco. All of them have flash arrays of one flavor or the other. Apart from IBM, none of them have anything analogous to a Cadence or a Resonance, or a Concert. EMC has Isilon, roughly analogous to Unison, though it has not faired well against Unison in recent competitions … we’ll be talking about this soon. IBM rid itself of things similar to the UCS platform. I wonder if Cisco will keep UCS. I think UCS is doing well, but its is competing against some very aggressive players in this space now. This competition looks to be interesting. I’d say that EMC, Netapp, and possibly Cisco have some more acquiring to do. they need big data appliances (gulp!) to compete with a leaner/meaner IBM now.