Just installed this. Big difference in temperatures. I noticed the fan speeds dropped right off. CPU temperature dropped 6 degrees, GPU dropped 7 degrees - it's almost quiet in here!
CPU performance has dropped a bit to around 88% from 100% - this is with 2 Aqua (CPU), 1 DistrRGen, 1 GPUGrid.
CPU performance goes up to 97% and temperatures go back up again once the GPU WU is suspended and Aqua WU is restarted - back being noisy again
Version 6.6.36
-
- Marvin the Dalek
- Posts: 4396
- Joined: Wed Mar 15, 2006 12:00 am
- Location: North Wales
Tried this but I've had to downgrade to 6.6.31.
I kept getting "Message from server: (won't finish in time) BOINC runs 63.7% of time, computation enabled 100.0% of that" on FreeHal despite the fact that Boinc runs 100% of the time and FreeHal takes virtually no CPU resource.
Apparently, it's something to do with Boinc averaging the time needed for all WUs currently downloaded and figuring out whether any work should be downloaded. Since each FreeHal WU takes 35 minutes, it seems to think that all cores are being fully used during that time.
I kept getting "Message from server: (won't finish in time) BOINC runs 63.7% of time, computation enabled 100.0% of that" on FreeHal despite the fact that Boinc runs 100% of the time and FreeHal takes virtually no CPU resource.
Apparently, it's something to do with Boinc averaging the time needed for all WUs currently downloaded and figuring out whether any work should be downloaded. Since each FreeHal WU takes 35 minutes, it seems to think that all cores are being fully used during that time.
Same worked for me (suspending WUs) but I don't think it's Aqua related.
It doesn't matter which WUs I suspended, so long as BOINC thought I needed more work, obviously Aqua work best as they take multiple days!
Only happened on the one machine that I upgraded to 6.6.36, going back to 6.6.31, everything's alright.
It doesn't matter which WUs I suspended, so long as BOINC thought I needed more work, obviously Aqua work best as they take multiple days!
Only happened on the one machine that I upgraded to 6.6.36, going back to 6.6.31, everything's alright.