Are we ignoring a whole lot of history ?
I don't know about Jerry Sanders but it appears he did one right move and that was NexGen (I recall being in AMD lobby and there was a poster of "Raiders of the lost Arc" but Jerry S. was Indiana Joe - very "funny" but what culture clash compared to Intel; you ask yourself how serious is this guy really - just think about AMD customer would visit the fab and see the poster);
At this time AMD's manufacturing sucked - the quarter was litterally hinging on keeping half a dozen metal etchers running 24x7; Duran (I guess the guy now in charge at Spansion) was in charge and he thanked us because we helped making the quarter.
All that is long time ago but it seems Hector with his Moto background (at that time Motorola' was pushing Six Sigma program) helped straighten out manufacturing at AMD.
While "googling" around I came up with a blurb that INTC's die size was 1.7x compared to AMD (2002) and supposedly at equal technology node AMD will always have an advantage with regards to die size. UMC was selected as manufacturing partner but it never took off.
Some of AMD posters are overly fanatic but it's a huge accomplishment to compete successfully with Fab 30 which is 200mm.
But one needs to look towards future:
1.) AMD needs to get 300 mm process going at 90nm to support manufacturing plan;
2.) at the same time must get 65nm going without impacting manufacturing
3.) my guess by the time AMD gets its 65nm Intel will start cranking up 45nm
4.) And than there is what I call "Pearl Harbour effect" - in case Intel underestimated AMD they now know what's at stake.
I could be wrong but I think it's overly optimistic that CHRT could pull off 65 nm way before AMD.
(BTW: was CHRT discussed during CC at all after all the positive noise)
I believe this was the one "wise" decision made by J.S.
cpu-info.com
BTW: looks like Intel pulled in additional third fab for 65 nm
IDF Taiwan: Intel confirms packages for Conroe and Woodcrest, but remains fuzzy on Merom
Printer friendly
Related stories
Comments
Email to a friend
Latest news
Related topic: Hot systems
Advertisement
Vyacheslav Sobolev, DigiTimes.com, Taipei [Tuesday 11 April 2006]
At IDF Taiwan, Intel officially confirmed packaging for two of its first 65nm dual-core processors (Conroe and Woodcrest) based on the next-generation Intel Core microarchitecture, but failed to clarify packaging on a third processor (Merom) using Intel Core.
According to Dan Casaletto, vice president of Intel’s digital enterprise group and director of microprocessor architecture and planning, the desktop Conroe processor will utilize the LGA 775 package, which has been in use since 2004 when it was introduced for the Pentium 4 platform.
The Woodcrest chip will target dual-processor servers and workstations, Intel said last month at IDF San Francisco. Casaletto confirmed that the processor is designed to support the new LGA 771 socket that will appear in the market with the coming launch of the 65nm Xeon processor known as Dempsey (Dempsey will be not based on the Intel Core microarchitecture).
Conroe and Woodcrest are both targeted to be launched in the third quarter of this year.
Casaletto also said that Intel intends to make its first quad-core offerings for desktops (Kentsfield) and dual-processor servers and workstations (Clovertown) pin-to-pin compatible with Conroe and Woodcrest, respectively.
Concerning the Merom processor for notebook applications, Intel was less clear. Casaletto stated Merom will initially support one of the current Core Duo and Core Solo processor sockets to maintain compatibility with the Napa notebook platform (both processors are available in 478-pin mPGA and 479-pin mBGA packages). And next year, it will move to another socket with the Santa Rosa platform.
However, Casaletto mentioned that a 479-pin mPGA package will be used to help Merom maintain compatibility with the Napa platform, but this does not seem to be currently supported by Intel’s Core Duo and Core Solo products, according to Intel specifications. During a later Q&A session, Intel representatives were not able to clear up the contradiction.
Intel already has three 12-inch fabs ramping up 65nm processor production (two in Oregon and one in Arizona), according to Casaletto. Last month at IDF in San Francisco, Intel mentioned just two of them as having commenced 65nm production, with a second Oregon fab being added since then. Later this year, Intel will add one more fab (in Ireland) to this list, according to the company. |
