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: hmaly who wrote (94249)2/19/2000 10:56:00 AM
From: Scumbria  Respond to of 1571040
 
hmaly,

Do you realize that Intel can write some special compiler taking advantage of this feature, which will create a world class benchmark and Elmer will be bugging us about this every other post for the next ten years.

I don't see how. The ALU does not have any direct software visibility, and it's only impact is that it effectively shortens the length of the pipe, while retaining the 1.5 GHz clock speed. This will speed up applications by 5-10% max.

Scumbria




To: hmaly who wrote (94249)2/19/2000 1:34:00 PM
From: Tenchusatsu  Read Replies (2) | Respond to of 1571040
 
hmaly, <Do you realize that Intel can write some special compiler taking advantage of this feature, which will create a world class benchmark and Elmer will be bugging us about this every other post for the next ten years.>

I think it's safe to say that even AMD will need some "special compiler support" for future processors. How do you think those x86-64 extensions are going to be supported? How about Sledgehammer's new RISC-like floating-point core which will no doubt need specific support from applications?

And like I said before, how many support is there right now for those "Athlon New Instructions"? AMD said that with a 266 MHz bus and "projected compiler and prefetch optimizations," the Athlon can experience a 40%-50% improvement in SPECfp_base95. When will that happen?

Tenchusatsu