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: KeithDust2000 who wrote (169726)	8/23/2005 4:46:34 PM
From: mcmab	Read Replies (1) \| Respond to of 275872

Performance/watt Well a lot of this comes down to what watts you're measuring. Given that there's a lot about turning off unused blocks within the processor I suggest that they are integrating the consumption over some time period. The watt then used would be an average figure rather than a peak. Thus to get a 5x increase in performance/watt is not difficult. It may amount to just a few % increase in performance AT a given frequency but a massive decrease in AVERAGE power consumption. That being said the door is also open for performance to be compared at a given frequency, say 2GHz, and NOT comparing the new architectures with the old Netburst. So how will the new processors, running at 2GHz compare to a Netburst at 3.8GHz?

To: KeithDust2000 who wrote (169726)	8/23/2005 5:25:22 PM
From: Rink	Read Replies (3) \| Respond to of 275872

Keith, mmm, I couldn't read the fine print at the bottom of that slide, but now you spelled it out it looks like that's what it says. Ok SpecIntRate. Still doesn't make sense. SpecIntRate improved from ~20 for Prescott to ~32 for Smithfield, a 60% increase. TDP increased from 115W for Prescott to 130W for Smithfield, a 13% increase. So that slide should have shown a 60-13=47% increase in SpecIntRate performance per watt, and it only shows a 20-25% increase. Maybe the difference isn't big enough to cry foul about especially as my SpecIntRate readings for Prescott and Smithfield are one out of more possibilities, mem. settings might be different, etc... I guess I simply can't figure Conroe at 5/1.4 * 65/130 = 79% better SpecIntRate which is 32*1.79 = 57. That would be incredibly good. What kind of X2 would it take to catch up with that only performance wise (not talking perf/watt)? Regards, Rink