Apple, AMD written off again

"APPLE IS IN long-term decline," observes Geek.com writer Sander Olson this week. "Even if Apple adopts the IBM PowerPC 970, no one at the Microprocessor Forum expressed any confidence in Apple's future. Every year more Macintosh users convert to Windows, and this trend will continue unless Apple comes out with another 'insanely great' product."

Macatistas will no doubt dismiss Olson's comment - and the comments made by MPF attendees, panelists and speakers that inspired it - as typical pro-Intel propaganda. AMD buffs will almost certainly level the same charge at Gartner Group analysts George Weiss and John Enck, who, just prior to Microprocessor Forum, doomed hybrid 32/64-bit chips to market failure.

Weiss and Enck were thinking particularly of AMD's upcoming Clawhammer CPU, but they could just as easily have been thinking of the IBM 970. Both chips combine native 32-bit operation for backward compatibility with advanced 64-bit processing for new applications. Unlike AMD, IBM is producing the 970 as much for its own use as anyone else's, but its most likely customer - and there can be few who doubt that it will buy the 970, or something like it - is Apple, and it has a lot in common with AMD. For a start, both claims that they're up the creek without a certain implement are bollocks.

Arguments that either company is doomed are based on the assumption that there's no mainstream market for 64-bit processing. That's broadly true, which is why good 32-bit performance is so important, but there are clear uses for 64-bit addressing beyond specialist scientific and large-scale database applications. Apple's digital media constituency will appreciate the 970's ability to support more than 4GB of memory because they can load even more high quality video footage into memory for editing and compressing to DVD, say. So too will AMD's small base of customers who build and sell workstations based on the Athlon MP probably have their eye on these kind of roles too.

Apple's financial future is predicated on serving a small, growing constituency of users. It isn't chasing massive market share gains, just trying to grow. Recession notwithstanding it seems to be making progress. Ignore the Windows switchers, Mac OS X is winning a positive response in the Unix arena, where it's the first version of that venerable OS to ship with a good-looking GUI (and no, I'm not forgetting NeXTStep).

AMD's financial future is also predicated on serving a small, but growing user base. It too knows it can't beat Intel (though never say never) but that it can build a good business on the back of folk willing to look for alternatives. To do that it has, like Apple, to deliver increasingly better products to its supporters, and attempt to come up with compelling reasons why others should come aboard too.

As I've said before, Clawhammer's 64-bit support is just such a reason. It probably won't win mass enterprise support, any more than Mac OS X will get Macs into big business, but like Apple's new OS (particularly when backed by the 970) it will give attract buyers who aren't the usual suspects as well many from among the faithful. Given both companies' business models, that's success.

Maximising that success will depend on execution, of course, specifically Apple's pricing and AMD's ability to deliver on schedule. Apple at least isn't dependent on others for an OS that can host all these whizzo new 64-bit apps. But AMD is able to sell into a much broader constituency of potential users than Apple can because it's x86 compatible.

One further similarity between Apple and AMD. If Apple does adopt the 970, it's unlikely to be able to ship machines before September next year. Clawhammer's latest delay puts its shipment in the same timeframe. I'll leave the (highly unlikely) possibility of collusion to conspiracy theorists, but it is going to make for an interesting time to have two key players both promoting 64-bit desktop computing at the same time. µ