Compiler problem while using GPU

1 次查看(过去 30 天)
khan
khan 2015-5-14
Hi, I am trying to use GPU for Matconvnet code on Matlab 2014b using windows 7. I installed Microsoft Visual C++ 2013 Professional compiler. Even it can be seen from below that it is perfectly working with mex codes (when i use cpu mode).
>>mex -setup
MEX configured to use 'Microsoft Visual C++ 2013 Professional' for C++ language compilation.
Warning: The MATLAB C and Fortran API has changed to support MATLAB
variables with more than 2^32-1 elements. In the near future
you will be required to update your code to utilize the
new API. You can find more information about this at:
http://www.mathworks.com/help/matlab/matlab_external/upgrading-mex-files-to-use-64-bit-api.html.
In CPU mode the code works. But when i go to GPU mode. It gives below error. Can any one guide whats the problem in it. Error is given Below.
vl_compilenn('enableGpu', true)
vl_compilenn: CUDA: MEX config file: 'C:\Program Files\MATLAB\R2014b\toolbox\distcomp\gpu\extern\src\mex\win64\mex_CUDA_win64.xml'
Starting parallel pool (parpool) using the 'local' profile ... connected to 2 workers.
Error using vl_compilenn>mex_compile (line 431)
No supported compiler or SDK was found. For options, visit
http://www.mathworks.com/support/compilers/R2014b/win64.html.
Error in vl_compilenn>(parfor body) (line 390)
mex_compile(opts, srcs{i}, toobj(bld_dir,srcs{i}), flags.mexcu) ;
Error in vl_compilenn (line 384)
parfor i = 1:numel(horzcat(lib_src, mex_src))
My GPU device is
ans = CUDADevice with properties:
Name: 'GeForce GT 640M LE'
Index: 1
ComputeCapability: '3.0'
SupportsDouble: 1
DriverVersion: 6
ToolkitVersion: 6
MaxThreadsPerBlock: 1024
MaxShmemPerBlock: 49152
MaxThreadBlockSize: [1024 1024 64]
MaxGridSize: [2.1475e+09 65535 65535]
SIMDWidth: 32
TotalMemory: 1.0737e+09
AvailableMemory: 971497472
MultiprocessorCount: 2
ClockRateKHz: 570000
ComputeMode: 'Default'
GPUOverlapsTransfers: 1
KernelExecutionTimeout: 1
CanMapHostMemory: 1
DeviceSupported: 1
DeviceSelected: 1
And i have Cuda 6.0 for it. How can i remove this error or fix it.
Regards khan
  3 个评论
khan
khan 2015-5-15
do you mean like this?
vl_compilenn('enableGpu', true) -v
Joss Knight
Joss Knight 2015-5-28
You need to edit MatConvNet's mex-compile function and add -v to the mex command. Just to be sure, you need to have installed the CUDA 6.0 toolkit, not just the driver. If you have installed it, maybe the bin directory is not on your system path - verbose output will tell you.

请先登录,再进行评论。

回答(0 个)

类别

Help CenterFile Exchange 中查找有关 GPU Computing 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by