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: Dan3 who wrote (94376)	2/20/2000 4:39:00 PM
From: Dan3	Read Replies (2) \| Respond to of 1571084

I know this has been brought up before, but upon further reflection, it might be reasonable to consider Willamette as a non-X86/X87 processor that can nevertheless emulate one - but only at much reduced performance, like Itanium. cpusite.examedia.nl Willamette will have a "weak" FPU. I've been looking through the documentation of Willamette last night (yes, I know, I'm addicted :-) ) and the latencies of the x87 instructions have been increased seriously. Multiplies and shifts are now much more costly. Also an instruction FXCH which is used to switch the contents of FPU registers is now not free anymore. In fact Intel recommends NOT using it anymore. To realize the seriousness of this you must understand the on the K6 family since CXT and the P6 family this instruction is free! Because this instruction can take away a lot of the downsides of the stack-based register model the x87 is still forced to use and because it was free it is used a lot. So existing code uses it a lot, and performance of those programs might be disappointing. Maybe Intel will do an Oracle - license the processor (instead of selling it outright) with benchmarking precluded except after review and edit by Intel. That way, only non-X86/X87 code (Willamette specific) benchmarks could be legally presented. Performance on the huge library of existing mathematical and scientific applications used by the workstation users initially targeted as users of this processor would be particularly poor. Dan

To: Dan3 who wrote (94376)	2/20/2000 11:52:00 PM
From: Paul Engel	Read Replies (2) \| Respond to of 1571084

AliBiDan - re: ". AMD called 1.1 GHZ 1.1 GHZ. Intel engineers presented 1.5 GHZ as 1.5 GHZ, but then had to watch some marketing types blather about parts of the chip running twice as fast - confusing the issue and casting suspicion over the whole demo" Intel reported exactly what they had - just as you said - a 1.5 GHz CPU with some internals running at 3 GHz. Confusion? Maybe. Confusion by people - SCUMBRIA NOTABLY - who REFUSED to accept that it was technologically possible. The confusion arose because Intel achieved that which has never before been done. An Innovation by Intel - which I guarantee you will be copied by AMD, IBM and every other CPU manufacturer within the next year or two. Jealousy and denial abound - and these are cast under the guise of confusion. Buckle up, AlibiDan - AMD is going for a Roller Coater ride as they try to keep up with Intel. Paul