vl_compilenn cannot find the nvcc executable in cudaRoot

[Anand Rajesh]

Hi I am trying to use 'vl_complienn' however the script is having trouble setting the command line path to point to the nvcc executable.

my command:

vl_compilenn('enableGpu', true, 'cudaRoot', '<CUDAROOTPATH>')

error:

Warning: Setting the 'MW_NVCC_PATH' environment variable to '/netopt/rhel7/versions/cuda/cuda-8.0/bin/nvcc' 

> In vl_compilenn>activate_nvcc (line 757)

  In vl_compilenn (line 296) 

Warning: NVCC not found in the command line path or the one found does not matches '/netopt/rhel7/versions/cuda/cuda-8.0/bin/nvcc'. 

> In vl_compilenn>activate_nvcc (line 768)

  In vl_compilenn (line 296) 

Error using vl_compilenn>activate_nvcc (line 780)

Unable to set the command line path to point to '/netopt/rhel7/versions/cuda/cuda-8.0/bin/nvcc' correctly.

Error in vl_compilenn (line 296)

  cuver = activate_nvcc(opts.nvccPath) ;

 

So what doesn't make sense to me is that I am unable to update my $PATH variable through matlab.

setenv(PATH) and getenv(PATH) are not changing the output of !echo $PATH or system('echo $PATH')


Any ideas on how to fix this? It seems like it is an environment issue.