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: Scumbria who wrote (59451)	5/24/1999 10:45:00 AM
From: Elmer	Read Replies (2) \| Respond to of 1573901

Re: "One of the fundamental principles of benchmarking is that- it is very easy to produce incorrectly low scores, but almost impossible to produce incorrectly high scores." Yes I realize this. But with release, according to you, only days away, you'd think someone would have a release version of the K7 and MB combo. Yet all we hear about is pre-release versions! Why is that? (Rhetorical question, no response required). EP

To: Scumbria who wrote (59451)	5/24/1999 2:50:00 PM
From: RDM	Read Replies (2) \| Respond to of 1573901

We have now reached new K7 development milestone: In order to launch a new CPU core we have to pass through key milestones: 1. Start design 2. Create tape out 3. Demo flacky prototype with visible limitations 4. Demo prototype in product format without visible limitations 5. Deliver samples to key people. 6. Leak benchmarks good and bad. 7. Announce product 8. Announce official company supported benchmarks 9. Deliver pilot product 10. Deliver full production (unsold inventory in stock) This thread is simply frothing about stage 6, above, without acknowledging that this activity is a typical and necessary precusor to stage 10. I tend to believe all of the stage 6. benchmarks tell something. However, knowing what they tell is something else and knowing conclusively without inside information is impossible. I tend to believe the benchmarks where the K7 provides better than expected results providing that it yielded the correct answers. Many benchmark programs do not check to see that the program did the work expected of it, i.e., it got the right answer. It the answers were correct it is unusual for a fast result to be wrong. Hence I tend to beleive the fast ones. Most defects cause slow results. I believe that the CPU most be working pretty well since there was not discussion of flacky or intermittent behavior. I would tend to expect the motherboard and support chips are responsible for some of the slow benchmarks. 8.