<div dir="ltr">Hi all,<div><br></div><div>Thanks a lot for your answers.<br><div><br></div><div>I followed Satish's advice and downloaded pets-dev. Then I tried to configure it with CUDA 6.0. It complained because I did not have flex installed, no big deal. Then it complained because it prefers GNU compilers than Intel ones, no big deal.</div>
<div><br></div><div>This being fixed, I was able to configure and compile pets-dev with CUDA 6.0. There was no problem due to deprecated arch or thrust directory.</div><div><br></div><div>I configured it with the following options:</div>
<div><br></div><div><div>--with-x=1 --with-mpi=0 --with-cc=gcc --with-cxx=g++ --with-clanguage=cxx --with-fc=gfortran --with-cuda=1 --with-cuda-dir=/usr/local/cuda-6.0 --with-cuda-arch=sm_35 --with-thrust=1 --with-thrust-dir=/usr/local/cuda-6.0/include/thrust --with-cusp=1 --with-cusp-dir=/usr/local/cuda-6.0/include/cusp --with-debugging=1 --with-scalar-type=real --with-precision=double --download-fblaslapack</div>
</div><div><br></div><div><br></div><div>Now this is done, how should I port my code to use it with CUDA ? Should I change something and include some CUDA directives in the code ? Is there some examples of makefile with nvcc that I could use ?</div>
<div><br></div><div>Thanks in advance.</div><div>Christophe</div><div><br></div><div><br></div><div><br></div><div class="gmail_extra">
<br><br><div class="gmail_quote">On Fri, Aug 29, 2014 at 6:29 PM, Jed Brown <span dir="ltr"><<a href="mailto:jed@jedbrown.org" target="_blank">jed@jedbrown.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="">Satish Balay <<a href="mailto:balay@mcs.anl.gov">balay@mcs.anl.gov</a>> writes:<br>
<br>
> On Fri, 29 Aug 2014, Dominic Meiser wrote:<br>
><br>
>> On 08/29/2014 08:31 AM, Matthew Knepley wrote:<br>
>> > On Fri, Aug 29, 2014 at 9:35 AM, Dominic Meiser <<a href="mailto:dmeiser@txcorp.com">dmeiser@txcorp.com</a><br>
>> > <mailto:<a href="mailto:dmeiser@txcorp.com">dmeiser@txcorp.com</a>>> wrote:<br>
><br>
>> > > Dominic, I think that thrust.py should depend on cuda.py. Do you<br>
>> > > know why it does not?<br>
>> > In principle you are right, thrust.py should depend on cuda.py.<br>
>> ><br>
>> > However, in my opinion, thrust.py should go away as a separate<br>
>> > package altogether. Thrust is shipped as part of any recent<br>
>> > version of the cuda toolkit (I forget since which version, Paul<br>
>> > might know) and it's always installed in<br>
>> > $CUDA_TOOLKIT_ROOT/include/thrust. Thus we can automatically<br>
>> > deduct the thrust location from the cuda location. Thrust should<br>
>> > be considered part of cuda.<br>
><br>
> I also think it should be removed.<br>
<br>
</div>Agreed. If it was correct to consolidate umfpack, cholmod, etc., into<br>
suitesparse, then there is no question that thrust should be<br>
consolidated into cuda.<br>
</blockquote></div><br></div></div></div>