nvidia drivers not responding

Questions and Answers : GPU applications : nvidia drivers not responding
Message board moderation

To post messages, you must log in.

AuthorMessage
Chris

Send message
Joined: 14 Jun 99
Posts: 12
Credit: 10,574,206
RAC: 6
United Kingdom
Message 844096 - Posted: 23 Dec 2008, 9:17:03 UTC

I've just returned to Seti after a few years lay-off and have started running a cuda version - with all latest drivers from nvidia. I've run a couple of hundred wu's pretty quickly without problems, but now it's getting computation errors on a high percentage of them. Not only that, but there's an error message at my end saying nvidia driver nvlddmkm is not responding, but the drivers have recovered. O no they haven't - my screen becomes mush, compounding with every wu that goes wrong. After three you can't make out anything. System reload cures it, and no obvious damage to anything. As far as I know I've got the latest everything as far as software is concerned.
H/w is GeForce 8600GTS. Anyone getting the same? What can I do?
ID: 844096 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 844105 - Posted: 23 Dec 2008, 10:31:27 UTC - in response to Message 844096.  

Read this thread to know what information I'd like and to recover, reboot the computer. I think your GPU is stuck, which can only be overcome by rebooting the computer.
ID: 844105 · Report as offensive
Chris

Send message
Joined: 14 Jun 99
Posts: 12
Credit: 10,574,206
RAC: 6
United Kingdom
Message 844301 - Posted: 23 Dec 2008, 23:00:49 UTC

Thanks for response. The system is Vista 32 bit Home Prem Sp1 (06.00.6001.00), the nvidia drivers are 7.15.11.7824 (7 Oct 2008). The cuda error reported is consistent over many tasks -

Cuda error 'find_pulse_kernel2<3, false>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1166 : unknown error.

The task gets aborted at that point.
Tasks invloved include 1097806722, 1098421568, 1091826354, and many others.

The GPU doesn't seem to be "stuck" as after the reported "recovery" of the nvidia drivers, following the error, everything carries on running. It's just that the individual pixels are not being displayed in the correct locations on the screen. This degenerates after each occurence of a cuda error and "recovery". Only about 10% are (seemingly randomly) wrong after the first error and recovery, but after the third virtually the whole screen is a dancing "mush" of random pixels.
ID: 844301 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 844320 - Posted: 23 Dec 2008, 23:19:37 UTC - in response to Message 844301.  

Cuda error 'find_pulse_kernel2<3, false>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1166 : unknown error.

OK, I have reported that one for Vista 32bit with 3 different driver versions to the developers already. Thanks for the report, though. :-)
ID: 844320 · Report as offensive
Profile enusbaum
Volunteer tester

Send message
Joined: 29 Apr 00
Posts: 15
Credit: 5,921,750
RAC: 0
United States
Message 844467 - Posted: 24 Dec 2008, 6:15:41 UTC

I'm having the exact same driver issue with Vista 32 bit.

I'm running the latest 180.84 drivers on an 8800GTX.
ID: 844467 · Report as offensive
Jörg

Send message
Joined: 10 Dec 02
Posts: 51
Credit: 1,547,286
RAC: 0
Germany
Message 844526 - Posted: 24 Dec 2008, 11:22:53 UTC - in response to Message 844467.  
Last modified: 24 Dec 2008, 11:44:41 UTC

Hello,

same problem here, I had several messages "Display driver did not respond and was restarted", the following WUs crashed too until I rebooted the PC:

Vista Ultimate 64bit
8800 GTS (512MB) - Driver 180.48

2 Examples:

WU 384603180

"Cuda error 'find_pulse_kernel2<4, true>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1172 : unknown error."

WU 4384172

"Cuda error 'find_pulse_kernel2<3, false>' in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_pulsefind.cu' in line 1166 : unknown error.
ID: 844526 · Report as offensive
Morten Ross
Volunteer tester
Avatar

Send message
Joined: 30 Apr 01
Posts: 183
Credit: 385,664,915
RAC: 0
Norway
Message 844970 - Posted: 25 Dec 2008, 12:23:41 UTC - in response to Message 844526.  
Last modified: 25 Dec 2008, 12:28:58 UTC

Hi,

I've got the same problem as well, and have also tested GPUGRID units in order to see if this is indeed a hardware issue or not. GPUGRID units are completed without a hitch, so hardware is not the case (some mention blown capacitors, or insufficient power/PSU etc.).

My system:

Vista x64 Ultimate SP2
Asus GeForce GTX 260
Cuda 180.84


Morten
Morten Ross
ID: 844970 · Report as offensive

Questions and Answers : GPU applications : nvidia drivers not responding


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.