Ok, spent most of the day palying around with my new rig and the gtx295 I bought
I have disabled SLI, and have tried using stock and optimised apps with no difference
The 295 is at stock clocks
I can only get one core crunching
The other core just produces errors after 19 secs, though I have seen some go through OK, but 99% of wu's error. For now I have disabled one of the cores.
A list of the errors
http://setiathome.berkeley.edu/results.php?hostid=5400786&offset=0&show_names=1&state=5
To summarise
<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
Work Unit Info:
...............
WU true angle range is : 2.289419
After app init: total GPU memory 939327488 free GPU memory 901906432
Cuda error 'find_pulse_kernel2<4, true>' in file '
</message>
Or I get a different error message
<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
Work Unit Info:
...............
WU true angle range is : 2.289419
After app init: total GPU memory 939327488 free GPU memory 901906432
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Have I got a faulty core, if so, how can I prove it. If not, anyone got any ideas on how to fix this?
I have disabled SLI, and have tried using stock and optimised apps with no difference
The 295 is at stock clocks
I can only get one core crunching
The other core just produces errors after 19 secs, though I have seen some go through OK, but 99% of wu's error. For now I have disabled one of the cores.A list of the errors
http://setiathome.berkeley.edu/results.php?hostid=5400786&offset=0&show_names=1&state=5
To summarise
<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
Work Unit Info:
...............
WU true angle range is : 2.289419
After app init: total GPU memory 939327488 free GPU memory 901906432
Cuda error 'find_pulse_kernel2<4, true>' in file '
</message>
Or I get a different error message
<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
Work Unit Info:
...............
WU true angle range is : 2.289419
After app init: total GPU memory 939327488 free GPU memory 901906432
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Cuda error 'find_pulse_kernel2<4, true>' in file 'd:/BoincSeti_Prog/sinbad_repositories/LunaticsUnited/SETI_CUDA_MB_exp/client/cuda/cudaAcc_pulsefind.cu' in line 1619 : unspecified launch failure.
Have I got a faulty core, if so, how can I prove it. If not, anyone got any ideas on how to fix this?