Apple and Intel

The speculation has been widespread since yesterdays late afternoon CNet mind bomb (deliberately timed to leave everyone twisting in the wind all weekend it seemed).

I'm sticking with what I had to say last week regarding the Apple x86 rumors. Local blogger Ole Eichhorn has an interesting take on things (I read it right before dashing out to finally see Star Wars — more on that later perhaps) which at least focuses on the developer side of the picture. Perhaps I've had too much of the Lucas vision today, but that's the dark side, the really painful aspect of this idea.

Matt Deatherage (editor and publisher of MDJ and MWJ), an extremely savvy Apple watcher and analyst, added this last night on the subscriber mailing list:

The entire point of the last issue was the emphasize that Apple could make products based around Pentium-family processors without "switching" the Macintosh to it, or even calling it a Macintosh. But the CNet story says that Apple is "scrapping its partnership with IBM and switching its computers to Intel's microprocessors."

The *only* way (so far) I can think of that this might make sense is if Intel gets into the PowerPC alliance, and is allowed to both manufacture and develop PowerPC-compatible processors. That would alleviate Apple's supply problem with IBM and help Intel legitimize a market that has struggled for a while. Everyone seems to be assuming that "Intel processors" means "X86-family," but there's no technical reason that must be true.

I hadn't previously considered the idea of Intel joining the AIM alliance originally formed by Apple, IBM and Motorola (is there really a PowerPC alliance or is that just what it's called these days?) and producing PowerPC silicon. Should that turn out to be the case, I think it would be terrific (competition is good).

In the end, if Apple does choose to use Intel x86 processors for future Macintosh products, it'll be interesting to watch how things turn out. As this article comparing the G5 and x86 platforms indicates, the much lower frequency Opteron continues to outperform the Xeon. Shifting processors also won't address the threading bottlenecks in MacOS X (see the chart here) and may actually make them more obvious.

Mondays keynote is going to have a lot more interest than it would have previously and while I haven't been to a WWDC in several years (and this one seemed like it was going to be especially boring), I'm sort of wishing that I could be there.