Tuesday, June 14, 2005

Apple's SEC filings and the Intel transition timeframe

If one searches through Apple's SEC filings, one can find details of the sales agreement between Apple and IBM. Agreement No. 000451 includes this statement:
This Custom Sales Agreement between Apple Computer, Inc. (“Apple”) and International Business Machines Corporation (“IBM”), which consists of this Base Agreement and Statement of Work Attachments, shall be referred to as the “Agreement”. The term of this Agreement commences on October 21, 2002 and expires on October 20, 2007.
This would explain much of Apple's reasoning for the two-year transition period to Intel, since both the Apple-IBM agreement and Apple's Intel transition period are slated to end in 2007. This works out well, since it gives Apple's engineers and third party developers enough time to get everything working properly and fast on Intel's chips. Furthermore, Intel hopes to have dual-core desktop/server chips with 64-bit support out by 2007 too, so the timing is perfect for a 2007 64-bit x86 Power Mac and Xserve update. Also, OS X 10.5 Leopard would already have been released by then, and should be fully 64-bit (unlike 10.4 Tiger). Interim Power Macs and Xserves would use new IBM chips (likely including the dual-core G5 970MP), while Apple's lower-end Macs would move Intel more quickly, using its low power 32-bit offerings (in 2006).

Some might wonder if Jobs had been chomping at the bit all along for such an x86 switch, with the partnership with IBM just being something to get Apple through the OS 9 -> OS X transition. However, I suspect if IBM had had more success with releasing faster G5 chips for the desktops and low power G5s for the laptops (and IBM had a promising future roadmap in terms of both chip technology and price) then Apple would have stuck by them.

Thanks to stingerman in the Ars forums for pointing me in the direction of the SEC filings.

Sunday, June 12, 2005

Ars posts hypothetical x86 Mac roadmap

Jon Stokes at Ars has posted his thoughts on the Apple-to-Intel transition and postulates a roadmap for Apple's adoption of x86 chips.

Much of it does make sense, but I'd be surprised if the Mac mini and iBook got the dual-core Yonah. The Mac mini and iBook are low cost machines, and I would expect them to get single-core variants of Yonah to keep costs down. The single-core Yonah parts would be more than sufficient, since even at 1.66 GHz, they are faster than the fastest G4 in current PowerBooks.

I would agree the PowerBooks will get dual-core Yonah chips. However, I am not entirely sure what will happen to the iMac. Like Ars, I would expect it to get a dual-core chip of some sort, but I don't think we can exclude just yet the possibility that it could get a single-core chip, since it is a consumer-oriented machine after all.