SI
SI
discoversearch

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.
Politics : Formerly About Advanced Micro Devices

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Yousef who wrote (27640)1/5/1998 3:56:00 PM
From: Petz  Read Replies (1) of 1572735
 
Yousef, Re: " If you get the higher frequency by increassing the current in the transistors (Idsat), then the power goes up as the square of frequency."

You say,

No John ... to increase the Idsat, you must either physically change the FET (shorter Leff, Gate thickness, Source/Drain resistance ...) or increase the voltage.

If you follow this thread back to where it started, increasing Vcc is exactly what I was talking about -- the K6 desktop CPU will run at a higher Vcc than the notebook CPU (in 0.25 um), therefore it will have a higher Idsat, higher operating frequency and power dissipation proportional to the square of the voltage ratio. For a particular CPU, the only thing you can play with to get a higher operating frequency is the Vcc (core voltage). This was confirmed by me recently for a K6-166 which operated flawlessly at 188 MHz at Vcc=spec level=2.9v, could do 200 and 208 MHz at 3.2v and 225 MHz at 3.3v. Similarly, its a sure bet that if K6 (0.25u process) notebook chips do 300 MHz at 1.8v, they'll do 350MHz at 2.1v. (K6-3D will go beyond that).

AMD's 0.25 generation has apparently physically shrunk the FET by more than the expected (0.25/0.35)^2 ratio, judging by the die size. This is why I believe AMD will catch up to Intel on CPU speed instead of being 2 speed grades behind.

Petz
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext