Author: Robert Hyatt
Date: 07:58:29 04/11/03
Go up one level in this thread
On April 11, 2003 at 03:06:56, Tom Kerrigan wrote: >On April 10, 2003 at 23:16:46, Robert Hyatt wrote: > >>On April 10, 2003 at 20:21:09, Tom Kerrigan wrote: >> >>>On April 10, 2003 at 17:07:29, Robert Hyatt wrote: >>> >>>>On April 10, 2003 at 14:51:11, Anthony Cozzie wrote: >>>> >>>>>>But I don't buy the 50% stuff, the cpu is not that simple internally. One >>>>>>thread will run at >>>>>>nearly full speed and the other gets slipped into the gaps, which is what I see >>>>>>(at least) when >>>>>>running Crafty. >>>>> >>>>>This statement seems counterintuitive to me. If one thread is given a higher >>>>>priority in the CPU, than the OS would need to take this into account (and not >>>>>just with PAUSE). Could you add a little debugging code to crafty and measure >>>>>how many nodes are computed by each thread on a run of your dual xeon? I'm very >>>>>curious if it is a 40/40/10/10 split or a 25/25/25/25 split (the latter seems >>>>>more logical to me). >>>>> >>>>>anthony >>>> >>>> >>>>I can't do it, because the threads are not bound to a logical processor, and >>>>they bounce around. >>> >>>Oh, think a little. >>> >>>Run two instances of single-thread Crafty and search positions with each. Then >>>you will see exactly how the processor's time is divided. I'll eat my hat if you >>>see 100%-10% instead of 55%-55%. >>> >>>-Tom >> >> >>Go look up Ingo Molnar's Linux test (he is one of the kernel developers and he >>wrote the original specifications for a SMT-aware kernel scheduler). >> >>His numbers bounced all over creation. 90-10 was as common as 50-50, which was >>why he got interested as I was asking questions and nobody had any answers... >> >>They were hardly _ever_ 50 50. They were usually _way_ off balance. > >I tried. Do you know how many hundreds of search results you get when you look >up Ingo Molnar and hyperthreading on Google? > >You'll have to give us something a _little_ less worthless than that to go on. > >And instead of referring to somebody else, why not just run a couple copies of >Crafty like I suggested? How long could it possibly take? A couple minutes? > >-Tom I have explained "why not" before. My configuration is a dual 2.8. I can't remove a CPU because I don't have a terminator to stick in the socket. So I am stuck with two. I can enable or disable SMT when I boot the machine. now tell me how to run the test. Two copies might run on one physical cpu (using two logical cpus). Or they might run on two physical cpus. I have no control over that. And they will bounce around between processors as they run. Your turn. Tell me how to run a valid test and I'll let 'er rip.
This page took 0.03 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.