SI
SI
discoversearch

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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?


To: eCo who wrote (71901)2/18/2002 10:10:00 PM
From: Joe NYCRespond to of 275872
 
Here is a portion of a good post on Raging Bull by economaniac:

Intel is ramping Northwood, and could potentially have large volumes of high performance processors. Shortages do not necessarily mean that Intel does not have decent yields on NW, but at least means they are not using them to crush AMD. Either way its good news for AMD. There is a limited window where Intel has .13 and AMD does not for Intel to really hurt AMD ASPs, if they survive that window unscathed that is very good for AMD, and (at least in a competitive sense) bad for Intel. This quarter is an Intel at bat. If nobody scores that's good for AMD and bad for Intel. If AMD does not start shipping tbred in volume in Q2 that an error that will let Intel extend the inning, otherwise the summer will see AMD try and take advantage of increasing capacity, access to new markets (sff and thin and light) and a richer mix (all benefits of their shrink to .13).

ragingbull.lycos.com

Joe



To: eCo who wrote (71901)2/18/2002 11:46:43 PM
From: Dan3Read Replies (1) | Respond to of 275872
 
Re: Low latency benefits the Athlon more than the P4. With the larger cache...

P4 is still stuck with an 8-way cache, while Athlon has a 20-way cache. P4's cache is simple, but fast, which is apparently needed to keep bubbles out of that long pipeline. With its limited set associativity, making the P4 cache much larger will be of limited use.

Maybe Intel can add some L3 with better associativity.