Marco
You should be able to use snow with open-mpi on your macpro. You should also be
able to use snow with the "sockets" method, e.g.
cl <- makeCluster(rep("localhost", 8), type = "SOCK")
Note that there's a separate list, R-sig-hpc, for parallelisation issues.
HTH
Richard
Blanchette, Marco wrote:
> I am fairly sure that this topic has been covered before. However, I can't
find answer to my questions in the archive...
>
> I have a very embarrassingly parallel computing tasks (we are doing ~13,000
anovas and ~48,000 t.tests) that could benefits from using the snow packages
parLapply() function.
>
> Does anyone knows whether I could use MPI or PVM to send jobs to the 8 cores
of my MacPro using snow or MPI/PVM has to be absolutely install and run across
multiple machine. Is there an alternative that I could use to take advantage of
my 8 cores with R?
>
> Thanks
>
> Marco
> --
> Marco Blanchette, Ph.D.
> Assistant Investigator
> Stowers Institute for Medical Research
> 1000 East 50th St.
>
> Kansas City, MO 64110
>
> Tel: 816-926-4071
> Cell: 816-726-8419
> Fax: 816-926-2018
>
> _______________________________________________
> R-SIG-Mac mailing list
> [email protected]
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>
--
Richard D Pearson [EMAIL PROTECTED]
Wellcome Trust Centre for Human Genetics http://www.well.ox.ac.uk/~rpearson
University of Oxford Tel: +44 (0)1865 287500
Roosevelt Drive Mob: +44 (0)7971 221181
Oxford OX3 7BN, UK Fax: +44 (0)1865 287501
_______________________________________________
R-SIG-Mac mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-mac
