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 236.78-6.1%Jan 30 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: pgerassi who wrote (79879)5/9/2002 10:07:21 PM
From: fyodor_Read Replies (1) of 275872
 
Pete: there is also an in between solution to expansion, each memory module has three ports. One towards the root, and a pair to leaves. Thus a single GPU HT link connects to one memory module which connects to two more, which connect to four more and so on.

The problem I see with that suggestion (disregarding any termination issues) is that the waste is quite large. If only a single module is used, 2/3 of the HT ports are unused. This would probably be the majority of systems, making the memory modules unnecessarily expensive. Even if all the slots are filled (three modules), 2/3 of the HT ports are unused on the two secondary modules.

In a commodity business, such as the DRAM business, I just don't think there is room for this kind of design inefficiency.

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