Replies to post #67320 on Advanced Micro Devices Inc (AMD)

[wbmw]

Re: the real question mark is the floating point performance, which Dothan is pathetic at

Dothan is really not so bad in floating point, except the really bandwidth intensive tests such as SPECfp, which are mostly held back by Dothan's limited 533MT/s FSB. Just look at how well the top bin Dothan performs in these two benchmarks:

http://www.gamepc.com/labs/view_content.asp?id=pm780&page=8

Clock for clock is does reasonable well vs. Athlon 64. Although, some people (on this thread) might ding the reviewer for using slightly lower latency 2.5 CAS DRAM for the Athlon 64, I'll let you be the judge on whether that totally disqualifies the results.... ;-)

[KeithDust2000]

alan, to be precise, their projection is:

A performance Dempsey system (2 Processors, 4 cores) will be 88% faster on SPECint_rate than a 3.6Ghz Irwindale system
(2 Processors, 2 cores). [performance level DP space H1/06]

A performance Woodcrest system (2 Processors, 4 cores) will be 52% faster on SPECint_rate than than the above mentioned Dempsey system. [performance level DP space H2/06]

So it´s a pretty detailed estimate for that benchmark.