Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Paul Engel who wrote (149110)	11/22/2001 10:42:26 AM
From: Dan3	Respond to of 186894

Re: what does this have to do with Intel's 2 GHz Pentium 4 built on their 0.18 micron aluminum P858 process - with 100 nM gates? It's easy to buy a chip, open it up, cross section it and measure the channel lengths. The outcome of such an exercise is that both Intel and AMD used .09 gates in their .18 processes when Intel was limited to 1.5GHZ in P4 and AMD was limited to 1.2GHZ in Thunderbird. Since then, it looks like both companies have moved to smaller gates in their .18 processes so that each gained an extra 33% in speed. Intel is putting up a brave face, claiming that Athlon is really a .13 chip now, while implying that P4 isn't (at the same time they trumpet beating AMD to .13 - they should pick a story and stick to it), but here's an independent 3rd party analysis of P4 and .18 copper Athlon from a company that dissects chips and sells the information they discover. Their dissections of .18 P4 and Athlon show the same gate lengths for both chips, despite the hopes of Intel fans. Note that, whatever gain Intel will be receiving from going to low-K has already been "used up" to make P4 while AMD can look forward to additional gains from low-k and/or SOI: Structural Analysis Report Intel® 1.5 GHz Pentium® 4 Processor Billed by Intel as the fastest desktop platform in the world, the Pentium 4 Processor achieves a 1.5 GHz operating speed through exciting process advances that include 0.09 micron transistor gates (as-drawn), a 0.18 micron process and low-k dielectric. The Pentium 4 Processor's nearest competitor, the AMD™ Athlon™ Processor, operates at speeds up to 1.2 GHz. It has a 0.18 micron process and gate lengths as small as 0.09 microns but uses 6-level copper metallization, not low-k dielectric. semiconductor.com Structural Analysis Report Now, I think it's become clear that .18 copper has an advantage over .18 Aluminum (despite Intel's squeals to the contrary), so Athlon does have that advantage, but then Athlon also has the advantage of SOI coming up, and, perhaps, low-k, so AMD can be expected to maintain a small but significant process advantage. Intel has already received whatever gains it will get from low-K - and the difficulty in polishing low-K material may explain what appear to be Intel yield problems. The "old" Athlon core clearly outperforms the P4 core. AMD has a new 64 bit core coming next year, with an extended pipeline to allow faster clock speeds - Intel won't have another major new core for 3 to 5 years. AMD's new core will have the marketing advantage of 64 bits - and that's got to be a very big advantage, especially with Intel pushing 64 bit Itanium, SUN pushing 64 bit Sparc, and IBM pushing 64 bit PowerPC. The only 32 bit chip remaining with any pretensions to performance will be P4, and 32 bit P4 will be competing with $25 32 bit Durons Fabbed by UMC. The indications are that P4's real performance won't be able to keep up with an SOI Hammer and that its marketing "performance" will have a tough time challenging a 64 bit CPU when it can only claim 32 bits.