Paul Hockett
unread,May 20, 2013, 4:30:03 PM5/20/13Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to lima...@googlegroups.com
Since the problems limapack deals with are massively parallel, I just wondered if you had considered heterogeneous methodologies? Both Nvidia and ATI/AMD have libraries for parallelisation across GPU workers which could be very nice in this case for desktop machines with recent GPUs (as opposed to more traditional cluster architectures). I assume the downside here would be that significant infrastructure may be required to get things running on but, optimistically, this could be quite simple to include and not so different from utilizing pthreads or mpi.