Many failed GPU workunits in a row
16 Apr 2016 19:30:47 UTC
Topic 198571
(moderation:
Several GPU workunits in a row failed, at least two dozen from various BOINC projects.
https://einsteinathome.org/task/555753121
https://einsteinathome.org/task/555753123
https://einsteinathome.org/task/555766226
https://einsteinathome.org/task/555766362
https://einsteinathome.org/task/555765887
https://einsteinathome.org/task/555765492
https://einsteinathome.org/task/555765494
https://einsteinathome.org/task/555765367
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.
Binary Radio Pulsar Search (Parkes PMPS XT) v1.57 (BRP6-Beta-cuda55)
windows_intelx86
Many failed GPU workunits in a row
)
I recently had the exact same problem.
I googled the error message "Cannot create a symbolic link in a registry key that already has subkeys or values." and the only thing I found was an old post about a nvidia driver issue where the GPU would "disappear" if the screen turned off as port of the power saving scheme.
Since the post was so old I was skeptical (but bugs have been re-introduced before), I changed the power saving scheme on the box to never turn off the screen (the box have no screen attached anyway) and the problem haven't happened after that.
Don't know if that fixed it or it was just a coincidence.
The OpenCL section of the
)
The OpenCL section of the Nvidia 364.72 driver, and earlier 364.* drivers, has a problem which can an entire computer to lock up, or cause a few dozen OpenCL tasks (often not all from the same BOINC project) to give a quick Compute Error. Problem not seen in the 362.00 driver.
Tasks from POEM@home seem the most likely to trigger this problem.
Threads on the problems:
https://www.primegrid.com/forum_thread.php?id=6769#94223
http://boinc.fzk.de/poem/forum_thread.php?id=1205#10896