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 221.06-1.1%3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: eCo who wrote (67442)1/9/2002 6:22:18 PM
From: Gopher BrokeRead Replies (1) of 275872
 
These SMT/SMP numbers are surprising. Look at the Dhrystone scores for the 1.8 GHz Prestonia.

All options enabled
2 x Prestonia .................8037 MIPS
SMT only (SMP disabled)
1 x Prestonia .................6688 MIPS
SMP and SMT disabled
1 x Prestonia .................3388 MIPS
SMP only (SMT disabled)
2 x Prestonia .................4425 MIPS


The Prestonia gains a staggering 97% (3388->6688) when SMT is enabled on a uniprocessor system but only 32% (3388->4425) for a dual SMP system where SMT is disabled. For that kind of increase when just enabling SMT, in the non SMT version the processor must be stalled for a huge percentage of the time (>>50%). And presumably the relatively small gain for the dual processor system is because the FSB is overloaded trying to feed a single processor so there is little benefit in adding a second.

But the regular XEON 1.6 shows a 100% gain when a second (presumably non SMT) processor is added? Something just doesn't look right with these numbers. Why would anyone buy a dual Prestonia Xeon when it only performs 20% faster than a single Prestonia Xeon (6688->8037)?

Anyone got any explanations for these strange numbers?
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext