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.
Politics : Formerly About Advanced Micro Devices

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Petz who wrote (40659)11/2/1998 1:12:00 PM
From: Tenchusatsu  Read Replies (1) of 1573954
 
<Ten, it would seem to be a good idea for Intel to make a server chip out of the Mendocino core. They could integrate 512K or 1M of half speed L3 (much cheaper than full speed), or full speed L3, into a Slot 1 or 2 cartridge. Wouldn't the reduction in bus traffic due to the internal 128K L2 improve 2 and 4 way server performance?>

What you are suggesting is adding a layer of on-die L1.5 cache between the L1 and L2 caches of Xeon. Unfortunately, this has no effect on front-side bus traffic, unless you also increase the L2 cache size. Adding the L1.5 cache only reduces back-side bus traffic. While this may increase performance somewhat, I'm sure Intel decided the ROI wasn't worth it, at least for the P6 core.

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