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: dougSF30 who wrote (239971)	9/7/2007 5:07:13 PM
From: Petz	Read Replies (1) \| Respond to of 275872

You are the one being ridiculous if you think cHT on K10 will be slower than K8. A bug in the initial stepping, sure, but if this were really serious, AMD would be releasing Phenom before Barcelona. As for the rest of it, once again, you seem to have forgotten all about Seaburg. Seaburg is a bandaid, and how will "improved snoop filter" help single-thread SPECfp? It won't. Intel refusing to reveal SPECfp at 3GHz is telling, isn't it? Since you don't believe me, here's the SPEC numbers with links at the end. I filtered to select cores/chip=4 and #of chips=2: bread and butter 2S Clovertown. Sort by frequency descending and baseline descending and you'll see I picked the fastest Clovertown systems == Intel/Supermicro submitted in Septemeber, except where no score was available and Fujitsu R540 was the fastest. MHz: 2000 2667 3000 scaling Tested By Platform SPECint 15.6(1) 20.8(2) 66.7% Intel Supermicro SPECfp 14.0(3) 16.6(4) n.a. 55.7% Fujitsu Celsius R540 intrate 92.2(5) 116 51.6%()Intel Supermicro fprate 56.8(7) 66.9 35.6% Intel Supermicro Links (1) spec.org (2) spec.org (3) spec.org (4) spec.org (5) spec.org (6) spec.org (7) spec.org (8) spec.org ()BTW, SPECint_rate scales by only 49% if Intel 10.0 compiler is used instead of v10.1 [2000MHz==>85, 3000MHz==>106]. Makes you wonder how well an ordinary server application scales compared to something coded in optimized assembly language. Well, I finally figured out why Intel took so long to reveal any 3GHz Clovertown benchmarks. Petz