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: Charles R who wrote (61191)	6/9/1999 9:12:00 PM
From: kapkan4u	Read Replies (1) \| Respond to of 1572611

<Chuck - re: Intel already covered its bases by hedging Rambus DRAM with PC100. Is it worthwhile for them to further segment the memory market (and increase system prices) if they are driving towards all-RAMBUS world in year 2000 or year 2001?> Chuck, I think that PC133 competes with Rambus for the future memory standard status. PC100 does not. Intel will do everything they can to prevent PC133 from gaining foothold. They will probably have a chipset that has a native support for PC133 (not a converter on Camino) by the end of this year, but they will not tell anybody that they are working on it. Kap.

To: Charles R who wrote (61191)	6/9/1999 10:47:00 PM
From: kash johal	Read Replies (2) \| Respond to of 1572611

Charles, Re: "What is your estimate of performance difference (at system level) between 800MHz Rambus DRAM, PC100 and PC133 memories when used in conjunction with Coppermine at 600MHz? I am not looking for precision here but a rough wag." I have heard and read that 300Mhz RDRAM(or 600 which is the same) is SLOWER than PC 133. I have also heard that 400Mhz RDRAM(800) is only marginally faster than PC 133 but for things like AGP 4x it does offer somewhat of a bigger boost. It is NOT clear to me that RDRAM buys anything over PC 133 in terms of SPEC int or FP numbers. Thats why I have been so critical of RDRAM and Intels descision to embrace it exclusively. IMHO Coppermine is a great processor as it should SLASH intels costs by 30% over current PIII. And their costs is therefore likely to be in the $40-50 range for a PGA or BGA package. But they are lumbering it with a more expensive chipset, much more expesive and hard to yield RDRAM. In fact no-one can even tell us what 256Mb of 800Mhz RDRAM will COST. It seems like they are fumbling and bumbling ala AMD. In this case the yield/manufacturabilty concern is related to the DRAM manufacturers and MB guys. And I agree that Cumine will be much lower performance than a K7. But the additional RAM/MB costs will push the OEMS costs to substantially greater than K7's. It could be a tough road to hoe!!! Regards, Kash

To: Charles R who wrote (61191)	6/9/1999 11:16:00 PM
From: fyo	Read Replies (2) \| Respond to of 1572611

Chuck: One thing that would probably tip the scales is if CuMine with PC133 would come close to K7 but that appears unlikely... It seems to me that people are underestimating CuMine somewhat. All indications are that it will be about 10% faster than a PIII in SPECint and SPECfp - based alone on the 256kB on-die L2 cache. Add to that a vastly improved chipset (including improved memory subsystem - be it using DRDRAM or PC133). This should further increase the performance. What improvements will there be in the K7 chipsets - or will the K7 be somewhat slowed down by an underperforming chipset? As for the MHz race... All indications are that Intel will have .18mu out several months before AMD. Surely this should go some ways towards eliminating the architectural advantage of the K7, no? At least until AMD can counter with a .18mu process of their own. (err... of Moto's, that is :)). --fyodor