[gmx-users] Re: cygwin_mpi_gmx installation

[라지브간디]

Thanks for your answers. I have uninstalled the mpi, have also reinstalled the CUDA and got the same issue. As you have mentioned before I noticed that it struggle to detect the CUDA.


Can cygwin recognize the CUDA installed in win 7? if so, how do i link them ?


The default installation path of CUDA in win 7 could be C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA


If I am correct, it should be like 


cmake .. -DGMX_GPU=ON -DCUDA_TOOLKIT_ROOT_DIR=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA


Kindly give me your suggestion. Thanks in advance.


Subject: Re: [gmx-users] cygwin_mpi_gmx installation 
To: Discussion list for GROMACS users <gmx-users at gromacs.org> 
Message-ID: 
<CANnYEw7Aq+strLpgC5nx7k9oENer5KOA--ajGUe6CN7zJRfVeA at mail.gmail.com> 
Content-Type: text/plain; charset=UTF-8 

Indeed it's strange. In fact, it seems that CUDA detection did not 
even run, there should be a message whether it found the toolkit or 
not just before the "Enabling native GPU acceleration" - and the 
enabling should not even happen without CUDA detected. 

Unrelated, but do you really need MPI with cygwin? Unless you are 
planning to run on multiple Windows machines, the default thread-MPI 
parallelization is enough to use multiple cores! 

Cheers, 
-- 
Szil��rd 


On Fri, Apr 12, 2013 at 1:04 PM, Mark Abraham <mark.j.abraham at gmail.com> wrote: 
> That looks really strange. CMake claims to detect CUDA but the right 
> variables are not set. 
> 
> 1) Can you reproduce this in a clean build directory? 
> 2) If so, what CUDA is present? 
> 
> Mark