After some fighting I managed to compile CUDA code on 64-bit Win7. I thought I had problems with my code, but after I checked code samples from Nvidia site I realised they don’t work for me too. I have tried two of them and they both seem to crash the same way. It is in file cutil_inline.h line 83:
max_gflops = device_properties.multiProcessorCount * device_properties.clockRate;
The stack:
0000000000831b70()
cudart.dll!00000001800189f7()
[Frames below may be incorrect and/or missing, no symbols loaded for cudart.dll]
cudart.dll!0000000180018e82()
cudart.dll!000000018001c8b1()
> matrixMul.exe!cutGetMaxGflopsDeviceId() Line 83 C++
matrixMul.exe!runTest(int argc=1, char * * argv=0x0000000000823310) Line 89 + 0x5 bytes C++
matrixMul.exe!main(int argc=1, char * * argv=0x0000000000823310) Line 77 C++
matrixMul.exe!__tmainCRTStartup() Line 266 + 0x19 bytes C
matrixMul.exe!mainCRTStartup() Line 182 C
kernel32.dll!000000007776f56d()
ntdll.dll!00000000779a3281()
My GPU is GT130M on Lenovo Y550 laptop. My OS is 64-bit Win7 Professional and I use VS2008.