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: andreas_wonisch who wrote (19041)	11/14/2000 1:34:44 PM
From: Pravin Kamdar	Respond to of 275872

andreas, Thanks for the P4 benchmarks. No doubt the Specfp2000 scores are from a highly optimized compiler, and that we won't see anything near a 70% performance gain in real world apps for quit some time. All the other marks look pretty bad. Intel needs the 1.6 at intro. Pravin.

To: andreas_wonisch who wrote (19041)	11/14/2000 1:55:38 PM
From: porn_start878	Read Replies (3) \| Respond to of 275872

Comparisons to a TB 1.2/266DDR : PIII-1000 PIV-1400 PIV-1500 Quake III – demo2 143.2 200.8 206.5 TB: 176 SYSmark J 1853 2013 2073 TB: 1916 (2000) and 2086(Win Me) amd.com SYSmark 200 200 208 TB : 245!!! Content Creation Winstone2000 53.5 58.8 60.2 TB: 63.5 SPECfp2000 314 541 562 TB : 370 anandtech.com 3D WinBench 2000 processor test 1.93 2.44 2.55 TB: 2.17 3DMark 2000 CPU speed test 393 434 457 TB: 434 sharkyextreme.com Max NOTE : the system configuration may differ but each one uses about the TOP of what is available...

To: andreas_wonisch who wrote (19041)	11/14/2000 1:58:13 PM
From: Pravin Kamdar	Read Replies (2) \| Respond to of 275872

Andreas, I like the last line of the Aussie review: But at the moment the new CPU just isn't much faster than established alternatives, and it's awfully expensive. Pravin.

To: andreas_wonisch who wrote (19041)	11/14/2000 2:48:49 PM
From: fyodor_	Respond to of 275872

<Andreas: More official and independent P4 benchmarks> Rant warning... ;) Putting out a new processor design is always a double-edged sword. On the one hand, it's necessary to do it in order to keep increasing the performance. On the other hand, all the existing code and compilers are optimized for your old design. If the new design differs significantly from your old, a lot of existing code (and new code for some time) will perform quite poorly on your new design. This was the case with the PPro (and even, to some extent, the P5) and will be the case for the P4. However, this rarely seems to matter all that much, except to the few folks benchmarking the processors and those ranting about the results of those same benchmarks. While it may take a bit of time for the optimized (non-benchmarking) code to propagate, it will happen. Especially if you supply 80% of all the processors ;). That said, AMD might very well have a secret card or two up its sleeve with the Palomino. There have been persistent rumors (coughJCcough) that it will have SSE. Providing the implementation is actually good, this would allow AMD to leverage a lot of Intel's optimizations. Combined with this "hardware prefetching thing" that everyone seems to be taking for a given, the Palomino could perform incredibly well on a variety of benchmarks. I certainly look forward to see the benchmarks. Both processors look to have real strengths, the P4 maybe a few more weaknesses. The question, of course, is whether the scalability advantage can beat the weaknesses (and if so, by how much?). Ok, end rant... -fyo

To: andreas_wonisch who wrote (19041)	11/15/2000 3:25:59 PM
From: Daniel Schuh	Read Replies (1) \| Respond to of 275872

Andreas, I note without comment that the url you gave, australianit.com.au^504,00.html, no longer seems to be working, though I'm sure I pulled it up when you first posted it. I must also note that the spec numbers are somewhat improved over the billybob numbers from ace's from a couple months ago, 544/562 vs. 502/525 for int/fp respectively. I'd put odds on compiler tweaks vs. hardware on that one too, but who can say? I haven't compared any of the others vs. previous leaks, are there any correlations? Cheers, Dan.