[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 6 Jun 2013 20:42:14 +0200
From: magnum <john.magnum@...hmail.com>
To: john-users@...ts.openwall.com
Subject: Re: Adding more "Nodes" to MPI Session
On 6 Jun, 2013, at 16:49 , Marc Brinkmann <marc.brinkmann@...il.com> wrote:
> Thanks magnum, for your fast answer.
> Could you please tell me the syntax for OpenMP?
>
> I'd like to run a OMP benchmark.
To build with OMP, uncomment (i.e. delete the '#') this line in Makefile:
#OMPFLAGS = -fopenmp
Then "make clean" and re-build.
> If it's fast enough, I'll restart the session.
If you are to restart it I'd recommend downloading latest bleeding-jumbo branch first. This will likely be compatible (resumable) with upcoming 1.8-jumbo-1.
> PS: if the CPU also supports opencl, couldn't i use it?
Yes, that can actually be almost 4x faster, depending on your CPU driver. It probably uses auto-vectorizing with SSE2 while the CPU format does not.
magnum
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
