You know I have to ask, but do you have any screen saver active?
Hi Jord,
I use the standard windows one and set it to blank the screen after 10 mins. Same on all machines in the fleet. Other BRP (cuda) tasks seem to have been getting through fine, just these two wierd ones that errored out.
I think it's safest not to run any screen saver when you run CUDA (or OpenCL, or CAL for that matter), no matter how strong your video card is or how much memory there is on it. A screen saver will at all times interfere with the running of the calculations being done on the GPU.
Depends now on what you understand is a screen saver. Is turning the monitor off by the OS power options a screen saver? Is turning the monitor off by the power button a screen saver? How far are you going to nitpick?
BRP cuda errors
)
You know I have to ask, but do you have any screen saver active?
RE: You know I have to ask,
)
Hi Jord,
I use the standard windows one and set it to blank the screen after 10 mins. Same on all machines in the fleet. Other BRP (cuda) tasks seem to have been getting through fine, just these two wierd ones that errored out.
BOINC blog
I think it's safest not to
)
I think it's safest not to run any screen saver when you run CUDA (or OpenCL, or CAL for that matter), no matter how strong your video card is or how much memory there is on it. A screen saver will at all times interfere with the running of the calculations being done on the GPU.
RE: A screen saver will at
)
Um, no. A blank screen does not interfere with the GPU.
Your statement is only for screen savers which put demands on the GPU.
RE: A blank screen does not
)
Depends now on what you understand is a screen saver. Is turning the monitor off by the OS power options a screen saver? Is turning the monitor off by the power button a screen saver? How far are you going to nitpick?
After searching, I stand
)
After searching, I stand corrected:
http://forums.nvidia.com/index.php?showtopic=71197
However, you might have been more specific about this Windows phenomenon.