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 213.50+6.2%Dec 19 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: combjelly who wrote (68927)1/25/2002 4:27:35 PM
From: TGPTNDRRead Replies (1) of 275872
 
CJ, Re: <For stuff that is heavily threaded, it should help.>

Yes. It always helps if you can prevent stalls or continue to work around them.

But it's not going to help the misprediction pipeline flush problem. It's going to make it substantially worse, IMO, for some important real world tasks.

There was an argument around a couple years ago that the compiler would be smart enough to handle the threading -- kind of like the VLIW argument.

LOL!

Any threading a compiler can figure out is going to be relatively trivial(IMO) -- and the vast majority of benefit of multi-threading comes when there are large waits involved(think OLTP) and don't normally help the kinds of things that maximize CPU usage ( not the benchmark standard style).

But you know all this, 'cause you're in the business.

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