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: Paul Engel who wrote (59498)	5/24/1999 4:47:00 PM
From: Kevin K. Spurway	Read Replies (2) \| Respond to of 1574001

Re: "Still feeling real good about that Kflop 7, are you ?" I'm glad YOU feel better. Now maybe you'll sleep nights instead of regaling us with your monster truck inspired idiocy. Kevin

To: Paul Engel who wrote (59498)	5/24/1999 5:14:00 PM
From: Cirruslvr	Read Replies (1) \| Respond to of 1574001

Paul - RE: "Still feeling real good about that Kflop 7, are you [Scumbria]?" Heheh. You already have me laughing! I will be laughing at Intel even harder after the benchmarks from a FINAL K7 are published. All that will matter is AMD ability to execute, and we know how that has been in the past... You KNOW you are afraid of how powerful the FINAL K7 will be. But that is OK. We are your friends so you don't have to admit it to us.

To: Paul Engel who wrote (59498)	5/24/1999 6:07:00 PM
From: Scumbria	Read Replies (2) \| Respond to of 1574001

Paul, The FPU dispatch/scheduler would have to be designed to be reprogrammable, routing FPU instructions to only functional pipes - and the K7 would have to have on-board FPU diagnostics to DETERMINE which pipes were functional, and then REPROGAM the Instruction decoders/FPU Dispatch/Schedulers. Amazing! And all I thought you would need was a configuration register bit which marks the FPU reservation station as unavailable. Scumbria

To: Paul Engel who wrote (59498)	5/24/1999 8:09:00 PM
From: Petz	Read Replies (4) \| Respond to of 1574001

Possibility for low FPU benchmarks. The FPUmark score is certainly anomalous since they are inconsistent with a fully pipelined FPU as opposed to Intel's partially pipelined FPU (Intel FPU has 2 cycle throughput for multiplies, AMD one cycle). Rather than one of the three FPU units not working, my guess is that one or more 80387 instructions are not working, but are being emulated in software. The firingsquad article made a vague reference to "exceptions." Certainly the processing of a CPU exception, followed by software emulation of the instruction would skew a benchmark by 50%, even if it only involves 1% of the code in the benchmark. This could even be a software problem that WIN98 will need a patch to enable all of the K7 performance. Did firingsquad reinstall WIN98 and wipe out any drivers AMD had set up? Petz