Well, for almost a month I had problems running CUDA on my system with one of my 2 GPUs refusing to do work. See thread http://setiathome.berkeley.edu/forum_thread.php?id=58516 for details on that. This ended on Jan 27 when for some strange reason it worked fine... until Feb 7 when the other GPU, the one that worked originally decided to stop doing work. This results in every unit given to it failing out at about 40 seconds. Here is an example WU message:

<core_client_version>6.10.18</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
setiathome_CUDA: Found 2 CUDA device(s):
Device 1 : GeForce 9800 GX2
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1500000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 16
Device 2 : GeForce 9800 GX2
totalGlobalMem = 536543232
sharedMemPerBlock = 16384
regsPerBlock = 8192
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1500000
totalConstMem = 65536
major = 1
minor = 1
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 16
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce 9800 GX2 is okay
SETI@home using CUDA accelerated device GeForce 9800 GX2
setiathome_enhanced 6.09 Visual Studio/Microsoft C++
libboinc: 6.3.22

Work Unit Info:
...............
WU true angle range is : 2.594313
Optimal function choices:
-----------------------------------------------------
name
-----------------------------------------------------
v_BaseLineSmooth (no other)
v_GetPowerSpectrum 0.00044 0.00000
v_ChirpData 0.01789 0.00000
v_Transpose4 0.01199 0.00000
FPU opt folding 0.00212 0.00000
CUFFT error in file 'c:/sw/gpgpu/seti/seti_boinc/client/cuda/cudaAcc_fft.cu' in line 62.

</stderr_txt>
]]>

As you can see from the previous thread, the errors are the same, the only change is which GPU is being given the work.

It has also come to my attention that GPU 0 isn't alone in it's refusal to work, but GPU 1 randomly throws it's own in. This is rare atm but still annoying.

How can I fix this situation?

---