Advanced Micro Devices - Moderated (AMD) Message Board

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)

AMD 217.53

+1.5%

Nov 28 9:30 AM EST

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: Joe NYC who wrote (7250)	9/1/2000 2:49:47 AM
From: THE WATSONYOUTH	Read Replies (2) of 275872

Notice the power requirements. They seem to be going up exponentially above 866 MHz, and they don't even list 1.13 GHz. (If I am interpreting Intel' s tables correctly) The original 1Ghz part from the cBO stepping was 33W. The 1GHz part from the approx 15% shrink process cCO is 26.1W @ 1.7V and the 1.13GHz part from the cCO which also bumps the voltage to 1.8V is 35.5W. There is a 36% power increase from 1.1GHz to 1.13GHz for a .1V Vcc increase. This is a perhaps a 9% larger increase than you would expect if the channel lengths of the 1.0GHz cCO part were the same as the 1.13GHz cCO part. Also, if they could achieve 1.13GHz by bumping the Vcc alone, then there would be many more of them since most 1GHz parts could then be jacked to 1.13GHz. Clearly, there are not many at all. So, this 1.13GHz part must, in addition, represent the extreme short end of the channel length distribution. In fact, it probably represents a channel length regime where the short channel effects are just beginning to rapidly degrade the threshhold voltage of the devices. This will also explain the larger power jump than expected and the the 133MHZ jump in performance for a .1V increase in Vcc. It also may explain the problem since it is always dangerous to operate where the Vt of the devices starts to rapidly roll off with channel length. In this regime, you become much more sensitive to (unavoidable) even small variations in channel length across the chip which can lead to some circuit limited yield loss. Perhaps this is where Intel is parked with these recalled parts. However, if they can identify the parts of the logic that are failing, it is possible to fix it by raising the Vt locally via a new threshold adjust mask. If they can still clock 1.13GHz, perhaps we will see more 1.13GHz parts in Oct. all pure speculation of course THE WATSONYOUTH

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next