Author |
Message |
|
cufft: ERROR: /root/cuda-stuff/sw/rel/gpgpu/toolkit/r2.2/cufft/src/execute.cu, line 1070
cufft: ERROR: CUFFT_EXEC_FAILED
cufft: ERROR: /root/cuda-stuff/sw/rel/gpgpu/toolkit/r2.2/cufft/src/cufft.cu, line 147
cufft: ERROR: CUFFT_EXEC_FAILED
ERROR: CPME_cufft.cu, line 104: cufftExecC2R (gridcalc3)
04:53:14 (14404): called boinc_finish
Anybody know what's up with this.
Most recent is THIS one.
Ubuntu v9.04 64b, reverted from 190.xx back to 185.18 and got this one. I'll put the 190.xx back in.
____________
- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it." |
|
|
MarkJ Volunteer moderator Volunteer tester Send message
Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level
Scientific publications
|
cufft: ERROR: /root/cuda-stuff/sw/rel/gpgpu/toolkit/r2.2/cufft/src/execute.cu, line 1070
cufft: ERROR: CUFFT_EXEC_FAILED
cufft: ERROR: /root/cuda-stuff/sw/rel/gpgpu/toolkit/r2.2/cufft/src/cufft.cu, line 147
cufft: ERROR: CUFFT_EXEC_FAILED
ERROR: CPME_cufft.cu, line 104: cufftExecC2R (gridcalc3)
04:53:14 (14404): called boinc_finish
Anybody know what's up with this.
Most recent is THIS one.
Ubuntu v9.04 64b, reverted from 190.xx back to 185.18 and got this one. I'll put the 190.xx back in.
I notice you have a GTX260. It could just be the dreaded GTX260 fft bug that most 260's seem to suffer from. It came with cuda 2.2 (and cuda 2.3 didn't fix it).
____________
BOINC blog |
|
|
|
Is there another thread on this or a fix? I've not heard of this but it certainly seemed to start when I came up off of the Debian/Ubuntu 180.44 drivers.
Most recent: http://www.gpugrid.net/result.php?resultid=1324001
ERROR: CPME_cufft.cu, line 11: cufftExecR2C (gridcalc1)
12:16:59 (8263): called boinc_finish
PS: Also, are the GTX-275s immune?
____________
- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it." |
|
|
MarkJ Volunteer moderator Volunteer tester Send message
Joined: 24 Dec 08 Posts: 738 Credit: 200,909,904 RAC: 0 Level
Scientific publications
|
Is there another thread on this or a fix? I've not heard of this but it certainly seemed to start when I came up off of the Debian/Ubuntu 180.44 drivers.
Most recent: http://www.gpugrid.net/result.php?resultid=1324001
ERROR: CPME_cufft.cu, line 11: cufftExecR2C (gridcalc1)
12:16:59 (8263): called boinc_finish
There are a few. This might be of use (its the project admins asking for ones that don't work).
PS: Also, are the GTX-275s immune?
Well my GTX275 seems to be okay. So are my GTX295's. As far as I know its the GTX260's specifically. We wait for nvidia to fix it. It seems that Seti and Einstein cuda apps don't have the problem.
____________
BOINC blog |
|
|
|
Running Collatz on it w/o problems now.
____________
- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it." |
|
|
|
Well 6+ months have gone by... Did nVidia ever address this or is it beyond fixing?
____________
- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it." |
|
|
|
Never mind... found the thread that says CUDA 3.0 fixes it.
http://www.gpugrid.net/forum_thread.php?id=1420&nowrap=true#16536
____________
- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it." |
|
|