Talk:Intel P6

From Wikipedia, the free encyclopedia

[edit] When does a microarchitecture stop being P6?

Merom/NGMA are not P6 architectures. Why is this in the article?

Perhaps because the person who put it there thought of the NGMA as being enough like the P6 architecture that it belonged here. I'm not sure what you can do to the P6 microarchitecture before you have something deserving of being called a new microarchitecture - does the PPro -> PII transition count? PII -> PIII? Is the Pentium M just a P6, or should it be considered to have a different microarchitecture? Guy Harris 23:29, 29 January 2006 (UTC)

I've yanked the Merom/NGMA discussion out; if somebody wants to give information about NGMA, they should do so on the NGMA page. Guy Harris 23:38, 29 January 2006 (UTC)

Perhaps the Pentium M microarchitecture is just a flavor of the P6 microarchitecture, but is the Intel Core microarchitecture just a flavor of P6? How much does Intel have to do in order to get something that counts as a new microarchitecture? Guy Harris 03:15, 11 December 2006 (UTC)

Intel tends to define their microarchitectures by the number of stages in their execution pipeline. For instance, all Netburst machines have a 20 stage execution pipeline, regardless of front-end configuration, cache, etc. More importantly, and this really only applies to the processes inside Intel, the microarchitecture changes when a design team creates a new machine more or less from scratch. Pentium III and M and all of the other P6 variants are still P6 because the design team started with a P6 back-end when making those machines. With all of that said, it really boils down to marketing. -Fritz 21:48, 15 April 2008 —Preceding unsigned comment added by 74.195.238.198 (talk)