Replies to post #67420 on Advanced Micro Devices Inc (AMD)

[wbmw]

Re: in terms of producing low power CPUs, that SOI has been more effective for AMD than strained silicon has been for Intel.

I would disagree. Intel has scaled Dothan on 90nm at the following power and frequencies:

1.2GHz @ 5.5W
1.6GHz @ 10W
2.1GHz @ 21W
2.26GHz @ 27W

Yonah at 65nm offers dual cores at the following power and frequencies:

ULV @ 9W (according to the Inq)
1.66GHz @ 15W
2.16GHz @ 31W

These are power-frequency curves that K8 can't get close to.

Re: In the 2nd half 2006 when AMD begins volume production at 65nm using SSOI, Intel will NEED NGA to be competitive

That's not exactly a profound conclusion. To continue to lead in this business you need a great design and a great process. Intel has been minus a good design for a few years, but they've been able to get this far by maintaining a leadership process.

Re: 40-70w dc opterons should compare nicely to 80+7 watt Woodcrest

That remains to be seen. You are making a lot of assumptions on where AMD can clock and how much they dissipate with a process that's still filled with many unknowns. I'm still expecting Woodcrest to clock at close to 3GHz with performance per clock that's greater than K8. Also, I'm guessing your "+7W" number is based on the memory controller. Just keep in mind that Intel systems only need one memory controller, so for a DP system, it's actually 80W+80W+7W. And I bet AMD will have to crank up their power limits just to reach competitive performance levels.

Re: IF Intel's NGA is late, or their products don't perform as "projected" it could get quite ugly for their shareholders............at this point, AMD's promises seem more credible than those of their competition.

Yes, if Intel messes up with their next generation cores, they will be screwed. But I don't see AMD as any more credible in delivering to their next generation promises. As we have seen many times before, past execution is not an indicator of future performance.

Although, at least Intel has demonstrated their next generation cores in actual systems, while we are yet to see ANYTHING from AMD's 65nm process.

[imho]

AMD follows the leader

http://www.eet.com/news/latest/showArticle.jhtml?articleID=174901200

Andy Wei, a member of the technical staff based at AMD Dresden, ...AMD and IBM last year used a dual-stress-liner approach for their 90-nm transistors, putting differently configured nitride capping layers on top of the NMOS and PMOS transistors. At the 65-nm node, the partners added an embedded SiGe layer similar to what Intel used at the 90-nm node. Wei said “a big boost came when we added the embedded SiGe” at the source and drain regions of the PMOS transistor.

highlandpk: at this point, AMD's promises seem more credible than those of their competition.

Very good observation.<g>

IMHO