| Tuesday, October 12, 2004 |
16:45 - POWER5 Talks
http://arstechnica.com/articles/paedia/cpu/POWER5.ars
|
(top)  |
Here's the interview (via Steven) that Ars Technica's John "Hannibal" Stokes conducted with Pratap Pattnaik, the project manager for IBM's POWER5 CPU, the successor to the chip that spawned the PPC970 (aka G5).
Lots of red meat for the microprocessor geeks out there; but of course the inevitable thrust of the questions was whether this puppy would ever show up in Macs.
As I fully expected, Pattnaik could not discuss a possible workstation-class derivative (read: Apple-oriented derivative) of the POWER5. He also made it clear that he is and has been focused on POWER5 servers only, and any hypothetical workstation-class derivative of the design would be for someone else to discuss.
Nonetheless, he was very explicit about on one thing: a single-core POWER5 derivative for the workstation market will not happen, because single-core for any market but laptops is pointless from here on out. The entire industry is going dual-core across all market segments, so there is no possibility that POWER5 will be stripped of one core in order to be sold as workstation chip. Any workstation-oriented derivative of POWER5 would be dual-core from the get-go.
I want to make clear, of course, that the above comments do not in any way constitute a statement from IBM that a dual-core POWER5 derivative is in the works. From Pattnaik's perspective, these comments amount to stating the obvious, i.e., no one who takes a realistic look at where the industry is going should expect a single-core POWER5 derivative, because everybody's doing dual-core.
Interesting perspective from inside the industry; given all the setbacks that traditional CPU design has run into over the past year or two, one wonders what path it'll all go down next...
|
|
Brian Tiemann
