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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Scumbria who wrote (93431)	2/16/2000 12:20:00 AM
From: kapkan4u	Read Replies (1) \| Respond to of 1570818

<What are they calling the ALU? A bit shifter? I can guarantee that the control logic for the integer unit is not running at 3 GHz. The latch delays and clock skew would make that impossible.> I think you may have something here. They showed a 20-stage pipeline with a single execution stage. There is not much work that can be done in one clock at 3GHz. Maybe it is just logical and add operations that are double-pumped. Kap

To: Scumbria who wrote (93431)	2/16/2000 12:49:00 AM
From: Process Boy	Read Replies (2) \| Respond to of 1570818

Scumbria - <That is incredibly slimy of them to try to create the impression of an air cooled 0.18u part running at 1.5 Ghz.> You just keep right on setting yourself up. You are currently generating bookmarks in my "Stupid" file like crazy. The demo parts were plain vanilla .18, with the possibility of CD skews, but I don't know either way. I GUARANTEE you the median fmax of Willamette is way above 1GHz, again on said plain vanilla .18 process. All the demos today were air cooled, as far as I know, and I don't believe they would need special cooling. PB

To: Scumbria who wrote (93431)	2/16/2000 12:52:00 AM
From: Saturn V	Read Replies (2) \| Respond to of 1570818

Ref : What are they calling the ALU? A bit shifter? I can guarantee that the control logic for the integer unit is not running at 3 GHz. The latch delays and clock skew would make that impossible. The integer performance may not be as high as some people might expect, but definitely better than you think. I am quoting the Willamette programers manual. "very low latency integer operations can be completed at a rate higher than the processor clock frequency." " Floating point operations on Willamette will achieve a higher level than P6 when scaled to the same clock frequency". Unfortunately Intel is not disclosing everything.I still think that the FPU runs at higher than the processor rate, but I will have to wait it out. You may deny the Willamette claims. But unfortunately for AMD , Willamette is here sooner than we all expected. And if the design verification looks OK, the Intel Fab Capacity will finally be unleashed. The Coppermine story will not be repeated. (Unless the Willamette die size is >2X Coppermine size).