Message boards :
Graphics cards (GPUs) :
75% tasks failed with EXIT_CHILD_FAILED
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 14 May 15 Posts: 1 Credit: 50,470,937 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]()
|
Hello, I'm running GPU grid on a couple of systems and recently a LOT of my tasks are failing. State: All (1274) · Valid (307) · Invalid (0) · Error (958) All on this host: http://www.gpugrid.net/show_host_detail.php?hostid=500099 Antivirus reports no specific activity, Windows is up-to-date, drivers are up-to-date, card are not manually overclocked. (other GPU projects work fine) |
|
Send message Joined: 9 Dec 08 Posts: 1006 Credit: 5,068,599 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() |
There is nothing obviously wrong. I'm afraid your card may be failing, or factory overclocked (not expert). |
|
Send message Joined: 12 Jul 17 Posts: 404 Credit: 17,408,899,587 RAC: 0 Level ![]() Scientific publications ![]() ![]()
|
The WU I looked at failed for 2 others as well before it completed. # Engine failed: Particle coordinate is nan {nan = not a number} I don't see anything. Sometimes you just need to reboot. I run Linux and I always set a 16 GB cache. Yours is only 2 MB but maybe Win10 dynamically sets it. Are you running more than one WU per GPU? I recommend only running one WU per GPU. I have an FX-6300 and it's fine. I like to leave a CPU thread open for overhead etc on all my computers. |
©2025 Universitat Pompeu Fabra