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: KeithDust2000 who wrote (156619)	4/17/2005 4:20:32 AM
From: pgerassi	Read Replies (1) \| Respond to of 275872

Dear Keith: According to 4 core SPECweb 99 results, 4 840s (Angstrom Titan 64 Classic) beat the fastest 4 way 3.0 GHz Xeon MP score (HP DL580), 8800 to 8500. Now if 4 1.4GHz Opterons beat 4 3GHz Xeons, then 2 DC Opteron 275s (2.2GHz) will easily beat 2 DC 4.8GHz Xeons much less 3.8GHz ones. And the 275s come out this month, not middle of next year. For some reason no one lately did 2 CPU Opteron tests the last ones were 2 years old on 246s, they were done with just 2 drives to Fujitsu Xeons with 14 and RHCA 2 vs HCA 3.2. I'd ignore the results given the enhancement of two years of software development. As to JVM, again apples and oranges in the tests, very hard to compare using same software. There is the IBM using Hurricane versus an old Tyan S2882 board, but not one using the much faster S2892 board with Nvidia Nforce4 Pro 2250/2200s. As to TPC-C when using the same chassis DL380 to DL385, Opteron 252 wins 71K to 67K. The 74K result uses 1 more Smart array controller and 2 more disks 16 vs 14. It is more than likely if the 385 had 16 disks, it would have added 14% more TPC pushing it to 81K. It does win in the TPC/$ however. And they aren't using 64 bit mode where it would get even further ahead. When doing 4 way, it is easily ahead of 3.6GHz even against Hurricane, a 2.2GHz HP DL585 wins, 105K vs 102K. 2.6GHz gets 130K. Assuming linear scaling (best case), a 3.8GHz would only do 107K, not even close to a 2.4GHz Opteron 850 at 115K. A 3GHz HP DL580 does 95K, but thats a NW Xeon. A 3.8GHz NW Xeon would likely do, assuming linear scaling, 120K about equal to a 2.46GHz Opteron, still no comparison to a 852. So a 3.8GHz Prescott Xeon with 4MB L2 on a 1066MHz FSB probably would not beat a 275 much less a 280. And with the SRQ, a 275 is likely to be faster than 2 248s as the small amount of public benchmarks show that we have seen. In a few weeks, we are likely to have a much better idea of how a 275 or better would compare to Dempsey. I suspect that Dempsey is going to have to be better than is stated to have much of a chance to be roughly equal. And we do not know what Socket M2(M1) may hold to trump it. Pete