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: dale_laroy who wrote (35715)	4/14/2001 12:02:24 PM
From: Pravin Kamdar	Read Replies (2) \| Respond to of 275872

Dale, "If performance critical applications drive CPU power above its artificially low 54.7 watt limit, the CPU is halted with a 50% duty cycle (alternating 2 microseconds on; 2 microseconds off) until it cools down." I wonder if CNet and/or ZDnet will have the guts to report this incredible story. If the main line press picks up this story and exposes Intel's P4 clock throttling under heavy load "feature" to their customer base, it could be catastrophic for Intel. Thermal throttling on the P4 will have to occur more and more as Intel raises the core frequency before going to 0.13u, as their thermal limit will be reached more quickly as the clock is increased. Now, having said all this, I have not seen any reports form hardware sites that have tested the P4 to the effect that they have seen sudden drops in benchmark performance as the chip is pushed. We need experimental results to verify that this indeed happens. Pravin.

To: dale_laroy who wrote (35715)	4/14/2001 12:05:11 PM
From: fyodor_	Read Replies (1) \| Respond to of 275872

Dale: What an absurd approach. If they would just trottle back the clock speed from 1.5 GHz to 1.0 GHz they could drop the voltage. Sounds rather PowerNow!-ish ;-) (or even SpeedStep-ish) Seriously, though... the only possible reason I can think of that Intel implemented such a primitive solution is that the problem was only discovered very late in the project cycle. Maybe it's fixed in the high pin-count Willamette (if such a chip is actually released), but certainly the problem would seem to "go away" (for a while, anyway) once .13mu production is reached. -fyo