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)
AMD 221.05+1.4%Dec 8 3:59 PM EST
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Scumbria who wrote (1037)7/18/2000 2:37:26 AM
From: minnow68 of 275872
 
Scumbria,

You wrote "I guess the L2 isn't as bad as had been previously described on this thread."

That was precisely my reaction.

You also wrote "I'm not sure I agree with their decision not to do the L2 tag compare in parallel with the L1"

Since the L2 is 16 way set associative, the power draw would be double that of the coppermine's lookup circuitry for far less benefit (since the L1 is 128K).

Of all the AMD patents that I've read about, the "way prediction" patent was the one that really interested me. I wonder what kind of L2 set hit rate one could get? I would love an L2 cache structure where we have the 2-4 most likely tag sets compared with the L1 and the others only on an L1 or "active" L2 miss.

Potentially we could get most of the power savings and get most of the benefit of lower L2 latency.

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

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2025 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News