<div dir="ltr"><div></div><div class="gmail_signature">On Wed, May 25, 2016 at 12:58 PM, Uliano Guerrini <span dir="ltr"><<a href="mailto:uliano.guerrini@unimi.it" target="_blank">uliano.guerrini@unimi.it</a>></span> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex">On OSX 11.5 I instelled cuda 7.5, compiled samples and when I run:<br>
<br>
./deviceQuery Starting...<br>
<br>
CUDA Device Query (Runtime API) version (CUDART static linking)<br>
<br>
Detected 1 CUDA Capable device(s)<br>
<br>
Device 0: "GeForce GTX 680"<br>
CUDA Driver Version / Runtime Version 7.5 / 7.5<br>
CUDA Capability Major/Minor version number: 3.0<br>
Total amount of global memory: 4096 MBytes (4294508544 bytes)<br>
...<br>
<br>
and so on, so I presume my CUDA installation is correct driver is ok and card is supported. Then I compiled from source using openMP enabled clang from macport (the one from apple is useless and gcc isn’t supported anymore by NVIDIA on OSX) configuring with:<br>
<br>
cmake ../gromacs-5.1.2 -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DGMX_GPU=on -DCMAKE_PREFIX_PATH=/opt/local -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda<br>
<br>
All goes apparently well, apart from:<br>
<br>
-- Could not detect NVIDIA GPUs<br>
<br>
and:<br>
<br>
-- Could NOT find NVML (missing: NVML_LIBRARY NVML_INCLUDE_DIR)<br>
<br>
which I presume I can ignore as on OSX platform NVIDIA doesn’t offer NVML and GPU detection is based on that library (I also read on the net of people safely ignoring these warnings).<br>
<br>
Nevertheless when I run gmx mdrun I get the following message:<br>
<br>
NOTE: Error occurred during GPU detection:<br>
CUDA driver version is insufficient for CUDA runtime version<br>
Can not use GPU acceleration, will fall back to CPU kernels.<br>
<br>
My limited abilities (just some printf here and there) led me to find that the call to cudaGetDeviceCount on line 602 of the file <a href="http://gpu_utils.cu" rel="noreferrer" target="_blank">gpu_utils.cu</a> within gromacs/gmxlib/gpu_utils of the untarred 5.1.2 gromacs version fails.<br>
<br>
Any suggestions?<br></blockquote><div><br></div><div>Your questions are not related to development, please post such questions to the users' list (CCd, please continue there).</div><div><br></div><div>Secondly, this is intriguing, could it be that you are using a newer runtime form a different path with GROMACS than with the deviceQuery SDK sample?</div><div><br></div><div>BTW, I do think this should work, but to the best of my knowledge Apple has no official hardware that ships with or can be upgraded to GTX 680, right?</div><div><br></div><div>Cheers,</div><div>--<br>Szilárd<br></div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-style:solid;border-left-color:rgb(204,204,204);padding-left:1ex">
<span class=""><font color="#888888"><br>
uliano<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
--<br>
Gromacs Developers mailing list<br>
<br>
* Please search the archive at <a href="http://www.gromacs.org/Support/Mailing_Lists/GMX-developers_List" rel="noreferrer" target="_blank">http://www.gromacs.org/Support/Mailing_Lists/GMX-developers_List</a> before posting!<br>
<br>
* Can't post? Read <a href="http://www.gromacs.org/Support/Mailing_Lists" rel="noreferrer" target="_blank">http://www.gromacs.org/Support/Mailing_Lists</a><br>
<br>
* For (un)subscribe requests visit<br>
<a href="https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-developers" rel="noreferrer" target="_blank">https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-developers</a> or send a mail to <a href="mailto:gmx-developers-request@gromacs.org">gmx-developers-request@gromacs.org</a>.</font></span></blockquote></div></div>