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.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Petz who wrote (13998)	10/15/2000 4:16:38 AM
From: dale_laroy	Respond to of 275872

It will be interesting to see if AMD does the same with Mustang/Palomino. Back when Intel started this, I wrote to AMD suggesting that they design Mustang/Palomino such that individual cache lines could be disabled. For Mustang, this would mean that chips masked for yield of 2MB/1MB/512K L2 cache (which might be physically laid out as 17 or more ways from which only 16 ways per set were kept to yield the 16-way final product) could have 4 additional ways per set disabled to enable AMD to salvage Mustang processors with 1.5MB/768K/384K of 12-way set associative L2 cache. For Palomino the same approach would result in a salvage processor with 192K of 12-way set associative L2 cache. Note: A Palomino processor with exclusive architecture cache and 192K of 12-way set associative L2 cache (320K total) would have a higher hit rate for both instructions and data than P-III (with 288K total cache in a non-exclusive architecture), and could be the variant sold at the 100/200 MHz bus speed at 75% of the cost of the P-III with its 288K of total cache in an 8-way associative L2 non-exclusive architecture). The full speed (133/266) MHz FSB Palomino processors with their full complement of 256K of 16-way set associative L2 cache could be sold at the same price as the same speed grade P-III. With PC2100 DDR SDRAM, even the 1.0 GHz salvage variant should clobber the performance of Intel's 1.0 GHz P-III with PC2600 DDR SDRAM, even assuming no enhancements, such as wider data paths between cache levels.

To: Petz who wrote (13998)	10/15/2000 9:50:58 AM
From: niceguy767	Read Replies (1) \| Respond to of 275872

John: Don't think much else matters, other than the Athlon, in their PC division at this point...Virtually everything in the PC division (maybe even $4 to $5 billion in revenues) will be Athlon based in Y2001...Difficult to exclude "financial performance" in evaluating AMD's Athlon performance,which when factored into the equation, strongly suggests that AMD execution and performance have been phenomenal by almost any reasonable benchmark...No doubt AMD has faced its roadblocks along the way, and no doubt some have proved impenetrable yet, but such is the nature of the industry...Once the smoke and mirrors are cleared away, it becomes obvious that AMD, who one year ago could only compete, and ineffectively, in the low end of the microprocessor market, have successfully brought to market, the Athlon, which is now housed in products for all ranges (low, medium, high), products that offer superior performance at an affordable price...and not unimportantly, deliver a healty boost to the bottom line...In the big picture, on a scale of 1 to 10, 10 being the highest, it might be difficult for even the naysayers in our midst to rank AMD below 8 based upon execution and performance over the past year, unless of course, one chooses not to factor in profitability during the Athy's ongoing evolution over this period!

To: Petz who wrote (13998)	10/15/2000 1:27:57 PM
From: Charles R	Read Replies (2) \| Respond to of 275872

Petz, <Who cares? > CrazyMan! I was responding to his claim that AMD delivered everything AMD promised last year. <The only product Intel has is Pentium III. > For some reason people seem to ignore things like chipsets, laptop CPUs and server CPUs. And, the basis of the discussion was NOT intel. It was if AMD delivered everything that was promised last year. The answer is "no". <If some of the transistors don't work, they sell it as Celeron.> How does this relate to the discussion? Chuck