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: Tenchusatsu who wrote (60163)6/1/1999 8:18:00 PM
From: Petz  Read Replies (1) of 1570969
 
Tenchusatsu, I'd expect the K7 FPU to be no more than 20% faster than P6 architecture in pure floating point code.
The FADD pipeline is equivalent on both CPU's (1 cycle throughput)
The FMUL pipeline is twice as fast on the K7 (1 cycle throughput vs.2)
The miscellaneous pipeline I believe is somewhat more capable on the K7, being capable of more than just loads and stores.

Since typical FP code has twice as many FADD's as FMUL's, the one cycle FMUL penalty should only occur 25% of the time. This should give K7 an 8% advantage. K7 will get somewhat smaller gains because 1)it has lower latency (4 cycles vs. five, critical when you need the last result to continue a calculation) and because of the better miscellaneous pipeline.

Petz
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