Friday, January 14, 2011

AMD, ARM, NVidia & The Mobile Space

AMD Markham Canada
I've been following AMD for awhile now. They've been a fascinating company to watch. It's not easy being Intel's only real competition in the processor space and while AMD has never made it look easy, they have managed to survive and even thrive at times.

The recent ouster of CEO Dirk Meyer was a surprise to most people. Meyer was instrumental in turning AMD around over the past few years and boards don't generally fire CEO's with that kind of track record.  If news reports can be believed the issue of contention was one I've discussed here in the past, mobile computing. Apparently AMD's board was not happy with Meyer's lack of enthusiasm to invest effort in this area.

Even relatively large businesses such as AMD have limited resources. Figuring out how to allocate those resources to best position a company for the future is the CEO's most important job. Intel is large and rich enough that they can afford to make a few mistakes and focus on more than one area. AMD doesn't have that kind of flexibility. Meyer apparently felt that their best bet for the foreseeable future was to focus on the server and desktop space. Mobile devices are showing signs of eating away at that space but nobody believes the market for 64bit X86 compatible processors is going dry up any time soon. It's easy to see why Meyer felt the way he did. AMD has been locked in a life or death battle with Intel for decades and diverting resources from that effort apparently seemed like a bad idea to him. Continuing to focus there was also smart given AMD's poor cash position and their success in the high end server market where margins are much higher.

At some point AMD was going to need to devote resources to the mobile space. Meyer wasn't ready, AMD's board was, Meyer is now out of work. On the one hand it's kind of impressive that a board would take this active a role in the running of a company. More often than not corporate boards seem to be content rubber stamping whatever the CEO wants to do. On the other hand though this may make it difficult for AMD to hire a CEO capable of leading the company through what will be a challenging transition.

It's likely that the Intel/NVidia settlement played a part in the boards decision as well.

Intel, with far superior resources has found it difficult to create a processor that can compete with the plethora of ARM variants out there in the mobile space so it's difficult to believe that AMD will do any better if they decide to go it alone.  One advantage AMD does over Intel is a superior GPU.  AMD could go the route that NVidia is with Denver and couple a licensed ARM processor core with their GPU technology. This would be a relatively quick way to get into the mobile space. That market isn't short on existing competition and NVidia already has a head start in this direction so the road in that direction is unlikely to be smooth or easy.

Ironically AMD had a product line called Imageon that they sold to Qualcomm in 2009 that paired an ATI GPU with an ARP processor.

Ideally AMD is going to need to bring additional resources into the picture through hiring of acquisition if they are serious about competing in the mobile space. The time will come when it makes sense to shift existing resources away from their traditional desktop and server markets but I don't think that time has come yet.

Image via Wikipedia
Enhanced by Zemanta

No comments:

Post a Comment