Silicon Investor (SI) -- The First Internet Community

STOCKTALK

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: Elmer who wrote (120392)	7/16/2000 11:02:05 PM
From: Scumbria	Read Replies (1) \| Respond to of 1580023

Elmer, The only known example of K7 optimized benchmarks is the new SPECFP2000 scores. I don't know anything about the "optimized code", but I doubt that AMD has very much resource to put into the massively complex task of developing a top notch SPEC optimizing compiler. Scumbria

To: Elmer who wrote (120392)	7/16/2000 11:41:15 PM
From: Epinephrine	Respond to of 1580023

<Scumbria: "We've seen scientific computational benchmarks posted on the thread which show Athlon to be 40% faster than PIII" Elmer: We were talking about optimized code Scumbria. The only known example of K7 optimized benchmarks is the new SPECFP2000 scores.> Elmer, Actually I was talking about cores, if you reread my first post to you on this subject I stated that comparing based on optimized code involves too much non-architecture related code optimization technology and techniques to base sweeping architectural claims such as yours on. But speaking of optimizations... What do you think of Willamette's Achilles heel? I can understand why you are hyping code and instruction set optimization since if that old Aces Hardware article on Willamette architecture is true it supports the conclusion that the K7 core is 42% faster than Willamette on theoretical maximum performance with legacy (unoptimized) code. Message 12938265 Message 12938666 Message 12939911 I don't have a link to the old article but my conclusions must have been reasonable since you agreed with them at the time. (last link) Whew, I guess the big gamble at Intel is high MHz (even at cost of per clock performance) and code optimization!! Regards, Epinephrine

To: Elmer who wrote (120392)	7/16/2000 11:58:37 PM
From: Epinephrine	Respond to of 1580023

Elmer, Here is a link to that old Aces Hardware article: aceshardware.com Interestingly it also describes the exact thing that I was mentioning to you earlier about how the K7 core can do an add and a multiply at the same time whereas the P6 core can only do one at a time. (inherently superior architecture regardless of specific implementation performance.) Regards, Epinephrine "The FPU has two functional units, which is less than the Athlon! One for FADD and FMUL, one for FSTORE and FLD. In other words, the Athlon can do theoretically one floating point addition and one multiplication per clockcycle, while the Pentium IV (and Coppermine [P6]) can only do one multiplication or one addition."