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 : Intel Corporation (INTC)
INTC 36.26+0.5%Dec 18 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Cirruslvr who wrote (107834)8/21/2000 3:24:46 AM
From: Scumbria  Read Replies (1) of 186894
 
Cirrus,

"L1 data cache may be small for a very good reason: high clock speeds."

This guy azillionmonkeys.com thinks its latency will be 2 clocks vs. PIII and Athlon's 3 clocks, so I guess that will help make up for the L1 data cache being comparatively small.


In an x86 pipeline, the absolute latency of the L1 cache is not very important. This is because almost every instruction requires at least one L1 read, so the pipeline can be kept full. What is important is the hit rate, which is where a small cache will suffer.

If the 8K 2-cycle data cache is correct, the Willy architects seem to have made a critical error which will impact IPC and frequency. Willy is starting to sound like another Intel boondoggle.

Scumbria
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext