Message boards : Number crunching : xtreme overclocking?
Author | Message |
---|---|
Plum Ugly Send message Joined: 3 Nov 05 Posts: 24 Credit: 2,005,763 RAC: 0 |
I have a twin xeon 1.6@3.2 that doesn't seem to be turning out points as fast as other slower systems.Does anyone know if extreme OCing effects/slows down the rosetta or boinc programs. The system in question doesn't run anything but rosetta.It's not had any errors or problems system wise.No lockups or random reboots. The computer in question is h4 if anyone wonders. |
Tern Send message Joined: 25 Oct 05 Posts: 576 Credit: 4,695,450 RAC: 5 |
I have a twin xeon 1.6@3.2 that doesn't seem to be turning out points as fast as other slower systems. Do you mean that your times per result are slower, or that your credit claims are lower? If it's slower times, do you have the id# of a comparable system to compare against? There is a php utility that can compare hosts and give average times, etc. - look at http://www.andyk.de/boinc/. If it's credit claims, well, see the "Code release and redundancy" thread. It's possible that your benchmarks are "too low". They're looking at ways to even out the credit claims. Does anyone know if extreme OCing effects/slows down the rosetta or boinc programs. My AMD 3700+ is "moderately" oc'd, I'm quite sure many here go much farther than I have. If you aren't getting computation errors on the results, you probably aren't too "extreme". Prime95 is always a good stress test to make sure all is really well, but without errors, I wouldn't expect that's the problem. The computer in question is h4 if anyone wonders. What is an h4? If that's the computer name, we can't see that, only you can... |
Plum Ugly Send message Joined: 3 Nov 05 Posts: 24 Credit: 2,005,763 RAC: 0 |
Found the problem.I reinstalled it a newer boinc client and didn't merge results.Prime runs good at 3.6 on it with air an I lowered it to 3.2 for heat purposes. Sorry, changed it back to where everybody can see. Hid them to aggrevate 1 person.:-) |
Message boards :
Number crunching :
xtreme overclocking?
©2024 University of Washington
https://www.bakerlab.org