Credibility and Roadmaps?, gimme a break! It's plain as day that Intel has been operating in reactionary mode for the last 3 years.
"Desktops don't need 64-bit until 2010, later modified to 2007."
"4 GHz at end of 2003." WHere is it?
The clear Intel roadmap was to gradually replace x86 with IPF. Then Opteron "happened" and Intel wisely tore up that roadmap.
Its absolutely CLEAR that Pentium M was NEVER intended to be anything else other than a notebook chip. Now the P4 roadmap is being shredded and replaced by P-M.
On Intel's dual-core roadmap, only an idiot would start at desktop and then move to servers. Surely you don't think Barrett drew that plan up 3 years ago, do you? No, it's obvious that AMD made the right roadmap decisions with having a huge inter-chip bandwidth and low-power SOI technology. So dual-core happened at 90nm for AMD because their roadmap was intelligent, but could not be done at 90nm for Intel, because their roadmap was faulty. So the Intel response has been to put both dies on one anemic FSB, design chipsets that will be used for 6 months, and force mobo makers and OEMs to invest in technology that will be thrown away at 65nm. As for desktop DC, the entire Intel dual-core line is a throwaway short term reaction to AMD's steady roadmap. No idiot would buy a dual core P4 at the 65nm node. To quote Maximum PC August 2005 on the Pentium D and Pentium EE chips: "Our theory is that Intel rushed the Pentium Extreme Edition out the door before the X2 came along, so it would have a brief moment in the sun before comparision between the two CPUs was possible."
Again, reaction, not roadmap.
Did Intel have x86-64 on the roadmap? No, again, reaction, not roadmap.
Petz |
