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 -- Ignore unavailable to you. Want to Upgrade?


To: Scumbria who wrote (59872)5/28/1999 1:01:00 AM
From: grok  Respond to of 1573924
 
Re: <Kap, "I don't think they need to go beyond the 9-issue superscalar."
I once did performance analysis on a 4-issue processor, and discovered that the second slot gets about 50% utilization, the third slot gets about 5%, and the fourth gets about 0.1% utilization.
Also, given that only about one instruction gets retired per cycle, your assertion seems pretty darn safe. Scumbria>

Yes, it is likely that K7 and Willamette will represent the end of the road complexity-wise for x86 since there is little that can be done at this point beyond technology scaling (which should continue for about 8 more years. The ISA just doesn't allow much more room for innovation.

However, IA-64 is a whole different story.



To: Scumbria who wrote (59872)5/28/1999 3:02:00 AM
From: kapkan4u  Read Replies (1) | Respond to of 1573924
 
<Scumbria - re: I once did performance analysis on a 4-issue processor, and discovered that the second slot gets about 50% utilization, the third slot gets about 5%, and the fourth gets about 0.1% utilization.>

I myself wonder how can architecture (or a compiler) consistently utilize 9 issue slots. Do you know what was behind this particular K7 design decision? May be someone else can comment.

Kap.