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: TimF who wrote (106781)	4/19/2000 11:59:00 AM
From: Tenchusatsu	Respond to of 1574199

Tim, come to think of it, if Cumine-128 really is a Cumine-256 with half the cache disabled, it would be easier to disable half the ways instead of half the sets. That could be a reason why the L2 cache on Cumine-128 is 4-way instead of 8-way. But Anand says that he has confirmation from "reliable sources inside Intel" that Cumine-128 is not just a castrated Cumine-256. The difference in die size might be miniscule (like 10-15%), but with Intel's huge scale of manufacturing, those savings make a big difference. Tenchusatsu

To: TimF who wrote (106781)	4/19/2000 12:28:00 PM
From: Ali Chen	Read Replies (1) \| Respond to of 1574199

Interesting info from the Anand article: 1. "Classic" Coppermine has CPUID=681 2. "New" Celeron has CPUID=683 At the same time, the Intel's PCN #904 says that the new revised Coppermines, Core stepping B-0, all have CPUID=683. Coincidence? Not! BTW, what has happened with CPUID=682 parts? Massive dumpsterisation under ProsessBoy supervision? BTW, according to my personal experience, all current Coppermines (681) seems to remain very buggy and cannot execute some legacy code at all, while previous "0.25u" did it perfectly. In addition, the cheap plastic package seems to have poor thermal conductivity (as all plastics do), so frequent jumps in chip temperature (due to thermal management under OSes like NT and Win2000) causes huge thermal and therefore mechanical gradients inside the package body. I do not think the Cumines can last long. Tony Viola, how about RAS? One word - Floppermines. Throw in such pearls of other fine engineering as bad SMbus timing on all Intel's boards with MTH (hence recall of CC820 boards), and it makes one to wonder for how longer this cardhouse with stand. Never expected such flops from a $25B company with huge R&D budget. Funny. Actually, I did some predictions few years ago, but did not really believed in them myself. Now it seems it is moving towards reality. Regards, - Ali