Message boards : Number crunching : outta here
Message board moderation
Author | Message |
---|---|
Send message Joined: 15 Oct 04 Posts: 1 Credit: 6,550 RAC: 0 |
got tired of CP locking up my system. boinc fine, seti not causeing problems, just cp. detached. |
Send message Joined: 17 Aug 04 Posts: 56 Credit: 63,814 RAC: 0 |
Yeah... the CPDN client definitely does something that most other DC projects don't and that my CPU doesn't like at all. My system withstood over 10 hours of prime95's torture test with no errors but locked up hard within a couple minutes of starting CPDN. I was able to finish my last model by underclocking my system to 200 MHz below stock speed. I think I might try a model on my other good box and see how it turns out there. Anyone have any clue as to what CPDN is doing that might cause this? Maybe if you take out the function call to crash_cpu()? :) |
Send message Joined: 7 Aug 04 Posts: 2187 Credit: 64,822,615 RAC: 5,275 |
> Maybe if you take out the function call to crash_cpu()? :) Toby, Not sure what's up with some configurations that are having problems. My last nine models have all completed without error, with two more that will finish in the next few days. The computers I'm running them on are pretty much dedicated to crunching, and only CPDN. They aren't overclocked but running tight memory timings. I certainly don't have an answer for you. |
Send message Joined: 5 Aug 04 Posts: 907 Credit: 299,864 RAC: 0 |
Sorry, it's just not for every system unfortunately. You have to bear in mind that this is a huge supercomputer codebase that we are trying to run on PCs, so it's not as simple to implement for all system types as other projects. |
Send message Joined: 23 Sep 04 Posts: 3 Credit: 143,109 RAC: 0 |
> Yeah... the CPDN client definitely does something that most other DC projects > don't and that my CPU doesn't like at all. My system withstood over 10 hours > of prime95's torture test with no errors but locked up hard within a couple > minutes of starting CPDN. I was able to finish my last model by underclocking > my system to 200 MHz below stock speed. I think I might try a model on my > other good box and see how it turns out there. Anyone have any clue as to > what CPDN is doing that might cause this? Maybe if you take out the function > call to crash_cpu()? :) > I also noticed my cpu temperature is higher with CPDN than with other boinc projects. Regards Hans |
Send message Joined: 17 Aug 04 Posts: 753 Credit: 9,804,700 RAC: 0 |
> > I also noticed my cpu temperature is higher with CPDN than with other boinc > projects. Yes. CPDN is a very demanding application, and there is advice <a href="http://www.climateprediction.net/board/viewforum.php?f=23">here</a> especially in the hardware maintenance announcement. The power supply and cooling on some machines is simply not up to the job, and dust etc makes things much worse. |
Send message Joined: 23 Sep 04 Posts: 1 Credit: 6,252 RAC: 0 |
I've had it with the computer slowing down and locking up. So I'm going to quit cp. My question is, does the work my computer has already done on its first (and soon to be only) project get lost if I simply quit right now? Or do I have to finish this run (which will be finished in about two weeks). Anyone? Also, how do you quit? Do you simply quite the cp client and forget about it? Or is there more to it? |
Send message Joined: 10 Aug 04 Posts: 94 Credit: 309,849 RAC: 0 |
> > > > > I also noticed my cpu temperature is higher with CPDN than with other > boinc > > projects. > > Yes. CPDN is a very demanding application, and there is advice <a> href="http://www.climateprediction.net/board/viewforum.php?f=23">here</a> > especially in the hardware maintenance announcement. The power supply and > cooling on some machines is simply not up to the job, and dust etc makes > things much worse. > CPDN really gives the CPU a work out. My 3.4 is liquid cooled so no problem, but the 3.2G HT P-IV and 2.8G HT P-IV use heat sinks. I have had to manually set the fan speeds to max just to keep the CPU at below 51c on the 3.2 and below 49c on the 2.8. Good housekeeping is also essential. I clean everything out about every month or so. I don't bother to overclock because of all the attendant headaches that go along with it. On the plus side, I haven't had to use the centeral room heating at all so far this year. I just run the fan and let the computers provide all the heat. :-) |
Send message Joined: 17 Aug 04 Posts: 753 Credit: 9,804,700 RAC: 0 |
> I've had it with the computer slowing down and locking up. So I'm going to > quit cp. My question is, does the work my computer has already done on its > first (and soon to be only) project get lost if I simply quit right now? Or do > I have to finish this run (which will be finished in about two weeks). > Anyone? Yes, you have to finish the run for the results to be useful. The intermediate and end of phase trickles return very little data to the server; most of the results are uploaded at the end of the run. > Also, how do you quit? Do you simply quite the cp client and forget about it? > Or is there more to it? At the moment there is no simple way of telling BOINC that you want to stop doing a project at the end of the current WU. It is discussed in <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=1282">this thread</a>. To detach right click on CPDN in the projects tab, but it takes effect immediately. |
Send message Joined: 11 Sep 04 Posts: 9 Credit: 321,368 RAC: 0 |
I have 5 AMDs ranging from an XP1800 up to an A64 system and all of them except the 1800+ are overclocked with very tight memory timings. I haven't had any of them crash or hang though. I am guessing you have some other problem that CPDN is just making apparent. All of my machines are home made. |
©2024 cpdn.org