Dave Hudak <dhu...@osc.edu> wrote on 03/16/2010 04:27:35 PM: > > I have downloaded X10 2.0.1 and have gotten it running for single > and multiple places on my Mac laptop using the PGAS runtime just to > make sure I can get some things working. > > Now, I'd like to get it running on our production cluster. It's > pretty standard - x86, Red Hat Linux, infiniband interconnect. The > concerns I have are: > 1. batch process control - we use Torque > 2. infiniband network utilization - probably not essential for > my goals, but nice to have. > > I looked at the application development page here: > http://x10.codehaus.org/X10+Application+Development > ...and it looks like the MPI version of X10RT would solve these > issues, as I've built OpenMPI v1.3.2 in a way that's properly > integrated with torque and IB. > > Do I need to build X10 from source to create the MPI X10RT?
You will need to build a small part of X10RT from source and then drop a few resulting libraries/metadata files into the binary X10 distribution. Let me go through the exercise of doing it with X10 2.0.2 today and I will post detailed instructions of exactly what to do on the web page. --dave ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ X10-users mailing list X10-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/x10-users