I have been getting the above error message for about a month now. It is a pity that so much time goes wasted. I am using BOINC version 6.10.17.

Me too unfortunately. It appears to be pretty widespread too

---

Perhaps the new application will reduce such occurances?

It would be very nice for a system administrator to look into the problem of persistent "error while computing" messages that waste the time of the GPU processors for a month now.

Task output (http://www.gpugrid.net/results.php?userid=26804):

1780413 1119768 26 Jan 2010 4:10:13 UTC 31 Jan 2010 4:10:13 UTC In progress --- --- --- --- Full-atom molecular dynamics v6.71 (cuda23)
1778938 1118860 25 Jan 2010 19:11:12 UTC 25 Jan 2010 19:49:41 UTC Error while computing 2,189.11 144.74 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1778935 1118857 25 Jan 2010 19:07:51 UTC 25 Jan 2010 19:11:12 UTC Error while computing 97.45 13.23 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1778780 1118807 25 Jan 2010 18:02:06 UTC 25 Jan 2010 18:12:58 UTC Error while computing 557.72 63.87 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1778771 1118799 25 Jan 2010 17:49:47 UTC 25 Jan 2010 18:02:06 UTC Error while computing 632.88 53.48 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1778265 1118404 25 Jan 2010 17:41:36 UTC 25 Jan 2010 17:49:47 UTC Error while computing 400.50 48.13 4,716.44 --- Full-atom molecular dynamics v6.71 (cuda23)
1777393 1117846 25 Jan 2010 10:11:02 UTC 25 Jan 2010 17:41:36 UTC Error while computing 26,939.01 2,986.42 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1776203 1117201 24 Jan 2010 23:16:21 UTC 24 Jan 2010 23:25:06 UTC Error while computing 428.32 37.22 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1776154 1117164 24 Jan 2010 23:25:06 UTC 24 Jan 2010 23:36:16 UTC Error while computing 578.47 43.98 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1776053 1117091 24 Jan 2010 23:07:05 UTC 24 Jan 2010 23:16:21 UTC Error while computing 467.23 58.73 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1775981 1117050 24 Jan 2010 22:00:12 UTC 24 Jan 2010 23:07:05 UTC Error while computing 3,916.94 432.42 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1775722 1116859 24 Jan 2010 20:02:50 UTC 24 Jan 2010 20:07:15 UTC Error while computing 154.12 15.43 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1775428 1116699 24 Jan 2010 20:07:15 UTC 24 Jan 2010 22:00:12 UTC Error while computing 6,674.22 305.93 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1774370 1116015 24 Jan 2010 10:33:48 UTC 24 Jan 2010 20:02:49 UTC Error while computing 25,193.69 1,700.55 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1772444 1114818 23 Jan 2010 20:26:21 UTC 23 Jan 2010 22:00:18 UTC Error while computing 5,532.26 420.69 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1772398 1114779 23 Jan 2010 22:00:18 UTC 23 Jan 2010 22:56:37 UTC Error while computing 3,292.15 406.34 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1772318 1114721 23 Jan 2010 22:56:37 UTC 24 Jan 2010 15:34:02 UTC Completed and validated 48,903.24 2,446.67 4,022.81 5,430.80 Full-atom molecular dynamics v6.71 (cuda23)
1772250 1114669 23 Jan 2010 19:08:23 UTC 23 Jan 2010 19:45:44 UTC Error while computing 2,125.20 219.91 3,977.21 --- Full-atom molecular dynamics v6.71 (cuda23)
1772247 1114668 23 Jan 2010 19:45:44 UTC 23 Jan 2010 20:26:21 UTC Error while computing 2,316.09 162.49 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)
1771966 1114526 23 Jan 2010 17:43:24 UTC 23 Jan 2010 19:07:03 UTC Error while computing 4,901.45 365.09 4,531.91 --- Full-atom molecular dynamics v6.71 (cuda23)

I'm not from the project, but I just looked through the first page of your results.

Of all the WUs that errored out, not a single one of them got a second error when it was sent out again to another computer. Every single result either is still in progress, or was successfully completed by another computer. At least one of those successful results was completed by the same type of GPU as you're running (GTX250).

While not conclusive, that's a strong indication that the problem is with your computer.

First of all, you might want to start a new thread in the "Graphics Cards (GPU) forum". Server and website is for website problems, so a lot of people who could give you good advice probably don't even look here.

That being said, here's my thoughts on what might be wrong or might help:

1) Try powering off your computer (not simply restarting). I don't think this is your problem, but it doesn't hurt to try.

2) Install the latest stable release of the boinc client (6.10.18).

3) Try a different driver version. Some people on some projects have had better results with, say, the 190 or 191 driver rather than 195.

4) Does your GPU crunch other projects successfully? It's possible it's defective. There are GPU video ram test programs out there; it might be worthwhile testing the memory. Sorry, I can't provide a link, but I'l sure if you search for it you can find it.

5) Power supply? It's possible that if you're running on a stock OEM power supply, it might be having trouble powering the GPU.

6) Heat. Overheating can cause problems. GPU-Z can tell you how hot the GPU is running if you don't already have a utility that does that. You'll need to know what the safe or normal temperature range for the GTX250 is. The 250 is very different than the card I have (GTX280), so I don't know how it's supposed to behave.

7) Overclocking. If you're overclocking either the CPU or the GPU, try putting both of them back to stock speeds and see what happens.

Mike

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

GreekCluster, check your private messages.