On Wed, 16 Feb 2005 at 8:42am, Sean Dilda wrote > I'm a little confused as to why you want this. If you have dual > processors with HT enabled, then a full load is four processes running, > which would give you a load average of 4.0. And since ganglia will > indicate that a load of 4.0 is 'full load', then it currently does the > right thing.
If all 4 of those processes are compute intensive, then they're going to run slower in parallel then they would in serial sets of 2. Those virtual processors do share physical hardware, after all. > So I guess my question is, how is 4 processes running on a dual proc > machine with HT not full load? And if you want full load to be 2 > processes running, why are you using HT? Some may run HT in the hopes that with 2 compute intensive tasks running, the "idle" virtual procs will take care of the bookkeeping and other minimal load stuff the kernel's doing all the time without slowing down the compute tasks. I thought of running that way, but (as we all know) users are easily confused and I didn't think the possible benefits worth the pain of trying to educate them. ;) -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
