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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Dan3 who wrote (159449)	2/20/2002 1:18:53 PM
From: Paul Engel	Respond to of 186894

Blow Boy Dan - Re: "Time for you to go back and review the Sledgehammer design. The brilliance of AMD's design team has come up with a "third way." The brilliance of MAD's design team hasn't yet produced a saleable Shamster product.

To: Dan3 who wrote (159449)	2/20/2002 1:32:02 PM
From: wanna_bmw	Respond to of 186894

Dan, Re: "Re: The only alternative is to have separate memory per processor, and that requires ccNUMA support >> Time for you to go back and review the Sledgehammer design. The brilliance of AMD's design team has come up with a "third way."" Nope. A 4-way Hammer system will supposedly have local memory on each CPU, which still has the topography of ccNUMA. Since most software doesn't support having such local memory, AMD is simply willing to "eat" the latency, as it broadcasts transactions to every node. This will account for a huge penalty. AMD wasn't "brilliant" in this approach - they are simply brute-forcing their way into it by taking advantage of all the bandwidth on the Hypertransport links. Of course, that means that most of the benefit of all that bandwidth is going to be lost due to AMD's implementation. wbmw