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: ted burton who wrote (35874)	4/16/2001 12:57:26 PM
From: Ali Chen	Read Replies (1) \| Respond to of 275872

"Go find us that hypothetical nasty app." users.ev1.net see also the final section of this page: benchtest.com "because they understand what you do not - that after spending the extra dollars, their system will be indistinguishable from other systems" Yeah, it is like a Ford Mustang base model - Ford quotes the engine power at the red line... Undistinguishable? Not!

To: ted burton who wrote (35874)	4/16/2001 1:32:46 PM
From: pgerassi	Read Replies (1) \| Respond to of 275872

Dear Ted Burton: I already gave you one, Scimark. It reaches nearly 90% to 95% of worst case because it is composed of real scientific number crunching tasks. Another client, Seti, reaches 85% to 95% depending on config of worst case. Seti doing this make for a very nasty case, due to it using hitherto idle cycles, that being those that would cool a CPU (assuming that the OS used places CPU in stop grant mode (futher reduces power during idle periods)). Any of the so called "garbage" clients will cause the system to not cool down (garbage refers to the fact they they use idle or wasted cycles (stuff going to a waste bin is usually garbage)). I still say that this occurs unless the designer either 40% overdesigned the thermal system or used the true numbers and not the ones published. Benchmarkers typically use a motherboard on a bench where ambient air is readily available. If you put it into a case where there are no case fans beyond the one in the PS, pack it full of high performance things like graphics, high speed disks, burners, and so forth, and stick it into a closet (believe me, unsuspecting owners and users do this to quiet a system, or get it out of the way. I speak from experience of typical users in real environments), you get worst case conditions quite easy. If these are done to Intel's published TDP of 54.7W, there would be no margin to handle the above when actual power reaches to 70W or even 60W. Remember the only criteria is that the chip reaches 72C (actually the critical temp for the circuit to start trottling which has some varience probably +/- 2.5% or more). If the environment gets to 72C, it will trottle no matter how good the HSF. AMD's Athlon could still be running at full speed if the HSF is really good since the minimum maximum chip temp is 95C, the same margin that the P4 gets using the same rated HSF at an ambient of 49C. In this situation, P4 would always bench like a 500MHz P3 in a normal environment, a chip that costs $500 less. Not good from a users point of view. Pete

To: ted burton who wrote (35874)	4/16/2001 10:35:48 PM
From: Mani1	Read Replies (1) \| Respond to of 275872

Ted re <<All the information was, and still is, made available to the designers. None of them will design the thermal solution to the max power, because they understand what you do not - that after spending the extra dollars, their system will be indistinguishable from other systems, except for the added weight.>> Ted, you are 100% correct. It is refreshing to have someone with some real knowledge in this field. Thermal solutions are not usually designed at 100% of the TDP, nor should they be. Mani