Replies to post #66884 on Advanced Micro Devices Inc (AMD)

[wbmw]

Re: and of course as I have mentioned the X2 just like all rev E cpus has a TDP dependant on its TCase max.

Intel does the same. The Pentium M TDP, for example, is assuming 100C degrees:

http://download.intel.com/design/mobile/datashts/30526202.pdf (page 60)

Re: 89W being the absolute worst case scenario and seemingly rarely occurring.

AMD's 89W limit should *never* occur, since applications should never draw within the built in margin for max current delivery, one of the components AMD uses to calculate their overly conservative number. Personally, I think Intel's TDP helps vendors more. Not only does Intel have to guarantee that the processor will function within specifications if thermal conditions are designed to the TDP, but designing it based on what actual application code will dissipate, as opposed to a theoretical maximum, can allow 3rd parties to build less expensive systems.

Perhaps AMD feels that the extra thermal validation lab isn't worth the ROI, but I think they are leaving some additional competitiveness on the table. At some point, I predict they will switch to a more liberal definition of TDP, while still including the max number for power delivery purposes.