Thanks, I confirm that it works. You're doing us a big favor with your repository of Python packages, so thanks again.
I threadjack a bit for a minor annoyance on the subject of verbosity that was up a while ago on the list but persists in 2011.1: Every kernel I compile says (reproduced on Win7): C:\Python\Python27\lib\site-packages\pycuda\compiler.py:122: UserWarning: The CUDA compiler suceeded, but said the following: kernel.cu kernel.cu tmpxft_0000062c_00000000-3_kernel.cudafe1.gpu tmpxft_0000062c_00000000-10_kernel.cudafe2.gpu It works perfectly, but why? And can I turn it off? Best, Jonas On Thu, Apr 28, 2011 at 10:32 AM, Christoph Gohlke <cgoh...@uci.edu> wrote: > The PyCUDA binary you use was built with CUDAPP_TRACE_CUDA defined. Try the > updated installer. > > Christoph > > > On 4/28/2011 12:09 AM, brett.br...@csiro.au wrote: >> >> Hi >> >> I seem to have successfully installed PyCUDA on a Windows 7 64bit >> machine, Python 2.7.1, CUDA 3.2, PyCUDA 2011.1 using instructions from >> Christoph Bussler “Windows 7 64-bit with Visual Studio Professional 2008 >> (Strictly Binary Versions)”. Problem is… it runs very verbosely, and >> hence, slowly. Below is an example output from running the demo.py >> script from the PyCUDA source. >> >> How can I switch off the verbose output? >> >> Thanks >> >> Regards >> >> Brett >> >> C:\Users\bry074>python >> >> P:\SAF\AICAS\GPU_code_testing\Copy_LinuxGPU\python_scripts\pycuda_demo\demo.py >> >> >> cuInit >> >> cuDeviceGetCount >> >> cuDeviceGet >> >> cuCtxCreate >> >> cuCtxGetDevice >> >> cuMemAlloc >> >> cuMemcpyHtoD >> >> cuCtxGetDevice >> >> cuDeviceComputeCapability >> >> cuModuleLoadDataEx >> >> cuModuleGetFunction >> >> cuFuncSetBlockShape (doublify) >> >> cuParamSetv (doublify) >> >> cuParamSetSize (doublify) >> >> cuLaunchGrid (doublify) >> >> cuMemcpyDtoH >> >> original array: >> >> [[-1.85515511 -0.61387902 -0.9409495 0.57415462] >> >> [ 2.04725552 -0.43586037 1.19255412 -0.46156505] >> >> [-0.83301312 0.76718497 -0.53578204 0.16794606] >> >> [-1.24282026 -0.25060421 -0.50485265 -1.50926399]] >> >> doubled with kernel: >> >> [[-3.71031022 -1.22775805 -1.881899 1.14830923] >> >> [ 4.09451103 -0.87172073 2.38510823 -0.9231301 ] >> >> [-1.66602623 1.53436995 -1.07156408 0.33589211] >> >> [-2.48564053 -0.50120842 -1.00970531 -3.01852798]] >> >> cuFuncSetBlockShape (doublify) >> >> cuMemAlloc >> >> cuParamSetv (doublify) >> >> cuParamSetSize (doublify) >> >> cuMemcpyHtoD >> >> cuLaunchGrid (doublify) >> >> cuCtxSynchronize >> >> cuMemcpyDtoH >> >> cuMemFree >> >> doubled with InOut: >> >> [[-3.71031022 -1.22775805 -1.881899 1.14830923] >> >> [ 4.09451103 -0.87172073 2.38510823 -0.9231301 ] >> >> [-1.66602623 1.53436995 -1.07156408 0.33589211] >> >> [-2.48564053 -0.50120842 -1.00970531 -3.01852798]] >> >> cuMemAlloc >> >> cuCtxGetDevice >> >> cuDeviceGetAttribute >> >> cuDeviceGetAttribute >> >> cuDeviceComputeCapability >> >> cuDeviceGetAttribute >> >> cuDeviceGetAttribute >> >> cuDeviceComputeCapability >> >> cuDeviceComputeCapability >> >> cuDeviceGetAttribute >> >> cuMemcpyHtoD >> >> cuMemFree >> >> cuMemAlloc >> >> cuCtxGetDevice >> >> cuCtxGetDevice >> >> cuDeviceComputeCapability >> >> cuModuleLoadDataEx >> >> cuModuleGetFunction >> >> cuFuncSetBlockShape (axpb) >> >> cuParamSetSize (axpb) >> >> cuFuncSetBlockShape (axpb) >> >> cuParamSetv (axpb) >> >> cuLaunchGrid (axpb) >> >> cuMemcpyDtoH >> >> cuMemFree >> >> original array: >> >> cuMemcpyDtoH >> >> [[-0.48802722 0.91386962 -0.81973875 0.19833829] >> >> [ 0.51856601 -0.09402688 -0.48339829 1.20150411] >> >> [-1.65700185 2.41741824 0.0643658 0.15416692] >> >> [ 0.3448593 1.04530859 0.03367852 1.69314694]] >> >> doubled with gpuarray: >> >> [[-0.97605443 1.82773924 -1.63947749 0.39667657] >> >> [ 1.03713202 -0.18805376 -0.96679658 2.40300822] >> >> [-3.31400371 4.83483648 0.12873159 0.30833384] >> >> [ 0.6897186 2.09061718 0.06735703 3.38629389]] >> >> cuCtxPopCurrent >> >> cuCtxPushCurrent >> >> cuModuleUnload >> >> cuCtxPopCurrent >> >> cuCtxPushCurrent >> >> cuMemFree >> >> cuCtxPopCurrent >> >> cuCtxPushCurrent >> >> cuModuleUnload >> >> cuCtxPopCurrent >> >> cuCtxPushCurrent >> >> cuCtxDetach >> >> >> >> _______________________________________________ >> PyCUDA mailing list >> PyCUDA@tiker.net >> http://lists.tiker.net/listinfo/pycuda > > _______________________________________________ > PyCUDA mailing list > PyCUDA@tiker.net > http://lists.tiker.net/listinfo/pycuda > _______________________________________________ PyCUDA mailing list PyCUDA@tiker.net http://lists.tiker.net/listinfo/pycuda
