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.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: wanna_bmw who wrote (51264)	8/15/2001 4:31:50 PM
From: AK2004	Read Replies (1) \| Respond to of 275872

wanna_bmw re: How many people are going to buy a brand new computer, only to install applications of 4 years ago? eh, hmmm, well, most of the commercial sector which is most of the intel sales? :-)) Regards -Albert ps you asked......

To: wanna_bmw who wrote (51264)	8/15/2001 4:51:37 PM
From: Petz	Read Replies (1) \| Respond to of 275872

Wanda, there wasn't a single benchmark in the entire suite that was constructed by Bapco. Here's what independent reviewer Anand Shimpi said about Bapco Sysmark 2001: SYSMark's Internet Contention Creation test is very taxing on a memory subsystem. A large part of the test score is computed from a run of Microsoft's Windows Media Encoder while performing other tasks alongside it. "A large part?" Do most people creating web pages today spend "a large part" of their time encoding MPEG? There are many MPEG encoding programs available but Bapco happened to pick the one with the worst ratio of Athlon performance to P4 performance. Why is that? Even that program runs faster on AMD vs. Intel, in the more accurate floating point mode, but Bapco runs it in low precision MMX mode. Why is that? Petz

To: wanna_bmw who wrote (51264)	8/15/2001 6:16:35 PM
From: pgerassi	Respond to of 275872

Wanna_bmw: Then use some real benchmarks. Anand's database test done on real data and real usage. Scimark test done to real routines as currently used by programs doing real work. They even optimize for platform. Anand's is closest to what a real customer would test a real system. Those old tests using old applications are just what you find in the corporate center. The retraining and redoing the emperical fixes and work arounds are very costly. They usually avoid them, if at all possible. Besides those older applications run faster than the newer versions on the exact same machine (they couldn't bloat them on the old hardware (it wasn't fast enough then)). Also why recompile an existing application that has all of the bugs worked out? There is a reason for the old adage, "If it ain't broke, don't fix it!" Disobey it at your peril! Also use some data communications benchmarks using the standard DSP algorithms (they include lots of shifts and rotates), where the P4's lack of a barrel shifter causes much loss in performance. The P3, Celeron, Athlons and Durons all do very well at these as the data rates are relatively low and the code and data fit into L1. No your obsolete software remarks are just trying to save P4 face on older still relevant code. Intel didn't optimize the benchmarks to the P4's peculiarities, just the P2, P3 and Celeron. Pete