Why does the SMP client fail to utilize much of the CPU time

Testing the new OS X/Linux client in SMP mode.

Re: Why does the SMP client fail to utilize much of the CPU time

Postby Viper666 » Fri Jul 11, 2008 1:31 pm

ThomasHorsten wrote:Thanks for that explanation Vijay. I'll keep the SMP client running, and look forward to the A2 being stable on my system too :)

In other news, after I got numactl to work on my machine, I see a very slight speed increase. I run two instances of the SMP client, one bound to node 0 (core 0+1) and it's RAM, the other bound to node 1 (core 2+3) and the RAM attached there. Before I had numactl working, it was about 29.5 minutes per frame, but with the binding in place, it's now down to 28.5 minutes. Not as much difference as I'd have expected, but then most of the work is probably done in the cache anyway.

Thomas

now add

chrt 4 before numactl

chrt 4 numactl>the rest this works well on 2.6 kernels for me

Higher numbers can be used but 4 seems to let everthing work well together, but see what works for you.

The new kernels are Tickless I could get into explaining that but Its not a topic for FAH
ImageImage
Powered by AMD and Open SUSe 11.2 ----The New servers have nothing to do with fixing the stats problem !
User avatar
Viper666
 
Posts: 223
Joined: Fri Dec 14, 2007 9:57 pm

Previous

Return to Intel Mac OS X & Linux SMP client v6.0

Who is online

Users browsing this forum: No registered users and 1 guest