Replies to post #67343 on Advanced Micro Devices Inc (AMD)

[Ixse]

Keith, (EDITED) Anandtech slides refered to SpecIntRate/Watt. Not just SpecIntRate by itself. Big difference. See Y-axe tags on slides 3 and 5 here: http://www.ANANDTECH.com/tradeshows/showdoc.aspx?i=2503&p=2

But you probably have a different source of information, right? I remember you saying something like that in the past. I know we talked about this before, but can't find the conclusion back (which was that SpecIntRate projections were impressive). Could you please remind me, I mean if it doesn't cost you too much time?

Regards,

Rink

[Dan3]

Re: will be 88% faster on SPECint_rate than a 3.6Ghz Irwindale system

At the same clock running the same .exe? Or at a "projected" 5ghz running code from a radically tweaked new version of Intel's SPEC specific compiler?

My wife's Jeep is 88% faster than my Cobra - in the snow.

More to the point, by the time it ships, how will it compare with what AMD is shipping at that time?

We're still wating for all those "projected" performances from the (by now) 7ghz P4s, 2.5ghz Dothans, and 2.5ghz Itaniums for which we were hinted "projected SPEC scores".

Meanwhile, the difficult to manufacture (too tough for Intel, at any rate) dual layer strain SOI parts that incorporate on die memory controller plus chip scale interconnects for I/O don't need to have their scores "projected" - they can be measured.

[Dan3]

Re: to be precise, their projection is:

As long as the Dempsey and Woodcrest run at 5ghz and SPEC is compiled for a predetermined data set by the their "improved" multi-pass data optimizing compiler?

And Irwindale is benched with a real compiler like GCC or VC++?