Many failed GPU workunits in a row
Message boards :
Number crunching :
Many failed GPU workunits in a row
Message board moderation
Author | Message |
---|---|
Send message Joined: 17 Aug 14 Posts: 49 Credit: 5,225,280 RAC: 0 |
A workunit that appears to be missing one of the files required to run on a GTX 560: http://asteroidsathome.net/boinc/result.php?resultid=112153989 Or possibly, the CC detection didn't work properly. Part of a large cluster of GPU workunits from various BOINC projects that failed about the same time, so at least one of them did not leave BOINC and the driver in a state where another GPU workunit would start properly. Period Search Application v101.12 (cuda55) |
Send message Joined: 17 Aug 14 Posts: 49 Credit: 5,225,280 RAC: 0 |
The problem appears to be in the 364.* series of Nvidia drivers, and most often triggered by a POEM@home task. Once the problem is triggered, a few dozen OpenCL tasks (not necessarily all from the same BOINC project) will give a quick Compute Error, or the whole computer can lock up. Threads on the problem: https://www.primegrid.com/forum_thread.php?id=6769#94223 http://boinc.fzk.de/poem/forum_thread.php?id=1205#10896 The 362.00 driver does not appear to cause this problem. |
Send message Joined: 12 Mar 15 Posts: 1 Credit: 2,370,720 RAC: 0 |
Last modified: 24 Apr 2016, 15:39:47 UTC i also had this Problem. first i thought it is a hardwareproblem on my side, but after i read this thread here i downgraded the Driver to 362.00. it seems it is not only a Problem with POEM@home, because it is not installed here. until now all gpu-tasks running normal without Errors. thx for the info about that Driver Problem ;) . Boinc runs here on: Intel i7-3770K + IntelHD4000 Android-Stick-ARM-Cotex-A17 Sony-Z5C-ARM-Cortex-A53/A57 Nvidia GT-630 / Nvidia GTX-750Ti |
Send message Joined: 17 Aug 14 Posts: 49 Credit: 5,225,280 RAC: 0 |
Last modified: 3 May 2016, 1:35:32 UTC The problem appears to be in the 364.* series of Nvidia drivers, and most often triggered by a POEM@home task. Once the problem is triggered, a few dozen OpenCL tasks (not necessarily all from the same BOINC project) will give a quick Compute Error, or the whole computer can lock up. The 365.10 driver is now available, but does not fix this problem. The problem has also been seen on PrimeGrid. |
Send message Joined: 18 Jan 13 Posts: 6 Credit: 7,433,013 RAC: 1,885 |
Last modified: 8 May 2016, 21:43:25 UTC I am guessing that My PC is also suffering from this malady. http://asteroidsathome.net/boinc/result.php?resultid=114760634 What is going wrong? I am up to date with my BOINC and Nvidia software (well I thought I was, but there is an update to Nvidia, but from what I have read here, it is not a fix.). But from reading these posts, I need to back-grade my driver. That causes other programs to freak as they know what the latest driver for my boards need to be and as such will then not start. SETI@home_BETA_Test was having intermittent errors, but not as consistent as Asteroids. |
Message boards :
Number crunching :
Many failed GPU workunits in a row