Seti@Home optimized science apps and information
Optimized Seti@Home apps => Windows => GPU crunching => Topic started by: RottenMutt on 06 Jan 2010, 05:39:26 pm
-
In Win7 4 bit or Vista 64 bit I can't get all my cuda processors to be availiable with SLI OFF (latest drivers, all). If i enable SLI, then the cuda application will give results with errors, and then blue screen.
x58 classified
i7-920
2X GTX295
i can't get someone to comple cuda64 bit app, i would like to see if that works.
-
...
i can't get someone to comple cuda64 bit app, i would like to see if that works.
Cuda application development is pretty much stalled while there are driver problems (which is probably what you're seeing, is my guess) and SDK problems. Eventually there will be a specific 64 bit host code build, but not likely in a hurry (at least no-one I know of is working on one.)
Slowdown & issues have been reported by some with the latest 195.62 drivers, you might want to try going back to 191.07.
-
i guess it isn't as easy as telling the compiler to make it 64 bit!
well here is the output from failed work unit.
<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 GTX 295
totalGlobalMem = 939524096
sharedMemPerBlock = 16384
regsPerBlock = 16384
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1242000
totalConstMem = 65536
major = 1
minor = 3
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 30
Device 2 : GeForce GTX 295
totalGlobalMem = 939524096
sharedMemPerBlock = 16384
regsPerBlock = 16384
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1242000
totalConstMem = 65536
major = 1
minor = 3
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 30
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 295 is okay
SETI@home using CUDA accelerated device GeForce GTX 295
V12 modification by Raistmer
Priority of worker thread rised successfully
Priority of process adjusted successfully
Total GPU memory 939524096 free GPU memory 812523520
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
Build features: Non-graphics CUDA VLAR autokill enabled FFTW USE_SSE x86
CPUID: Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
Cache: L1=64K L2=256K
CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3
libboinc: 6.3.22
Work Unit Info:
...............
WU true angle range is : 0.601080
After app init: total GPU memory 939524096 free GPU memory 778969088
setiathome_CUDA: Found 2 CUDA device(s):
Device 1 : GeForce GTX 295
totalGlobalMem = 939524096
sharedMemPerBlock = 16384
regsPerBlock = 16384
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1242000
totalConstMem = 65536
major = 1
minor = 3
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 30
Device 2 : GeForce GTX 295
totalGlobalMem = 939524096
sharedMemPerBlock = 16384
regsPerBlock = 16384
warpSize = 32
memPitch = 262144
maxThreadsPerBlock = 512
clockRate = 1242000
totalConstMem = 65536
major = 1
minor = 3
textureAlignment = 256
deviceOverlap = 1
multiProcessorCount = 30
setiathome_CUDA: CUDA Device 1 specified, checking...
Device 1: GeForce GTX 295 is okay
SETI@home using CUDA accelerated device GeForce GTX 295
V12 modification by Raistmer
Priority of worker thread rised successfully
Priority of process adjusted successfully
Total GPU memory 939524096 free GPU memory 813236224
setiathome_enhanced 6.02 Visual Studio/Microsoft C++
Build features: Non-graphics CUDA VLAR autokill enabled FFTW USE_SSE x86
CPUID: Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz
Cache: L1=64K L2=256K
CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3
libboinc: 6.3.22
Work Unit Info:
...............
WU true angle range is : 0.601080
After app init: total GPU memory 939524096 free GPU memory 779681792
setiathome_CUDA: Found 4 CUDA device(s):
Device 1 : GeForce GTX 295
totalGlobalMem =
</stderr_txt>
]]>
-
i guess it isn't as easy as telling the compiler to make it 64 bit!
Yes and no ;) Can introduce even more unexpected problems fairly readily by ignoring problems before moving on, adding more complexity through x86->x64 portability issues... clearly the driver / cuda subsystem is having trouble there, and those information calls are functions of the driver.
That *should* work, and AFAIK does for many. 191.07 drivers + cuda 2.3 DLLs from the downloads section, I suspect don't have as much 'experimental' code designed for unreleased hardware.
-
i guess it isn't as easy as telling the compiler to make it 64 bit!
For CUDA - no, it isn't. When I tried to make x64 CUDA MB build (was no speed improvement so was not released) I had to uninstall x86 CUDA SDK completely and install x64 version.
Simultaneous usage was not supported. Hardly something changed in CUDA 3.0 in this.
Your stderr looks like task restarted few times from very beginning, no single checkpoint was made. But no errors in log...