SI
SI
discoversearch

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)
AMD 214.18-0.5%3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Joe NYC who wrote (69566)1/30/2002 6:17:14 PM
From: TenchusatsuRead Replies (4) of 275872
 
Joe, actually for a 4-way Hammer system, the latency is worse than you think. Each time a memory access is initiated, the other processors need to be probed. The access cannot be retired until all of the probes are resolved.

AMD's foils show a ring topology for a 4-way system. This means two hops for a probe to get to the furthest processor, then two hops for the probe response. This leads to a total of four hops, which should be considered the typical latency of any memory access.

It's curious why AMD would choose a ring topology instead of a mesh like IBM's POWER4. But then again, I believe 4-way and 8-way Hammer will be cancelled anyway. ;-)

Tenchusatsu
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext