[R-sig-hpc] multicore performance question

[Rainer M Krug]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 09/07/12 12:18, Mauricio Zambrano-Bigiarini wrote:
> I apologise. In my previous e-mail I did not write the time spent in each function:
> 
> # No multicore nr <- 500 ; X <- matrix(rnorm(1000), ncol=30, nrow=nr) set.seed(100) 
> system.time( unlist(lapply(1:nr, FUN=fn1, x=X)) ) #   user  system elapsed #  0.013   0.000
> 0.013
> 
> # multicore nr <- 500 ; X <- matrix(rnorm(1000), ncol=30, nrow=nr) set.seed(100) system.time( 
> unlist(mclapply(1:nr, FUN=fn1, x=X, mc.cores=6)) ) #   user  system elapsed #  0.030   0.380
> 0.125
> 
> Running on RHEL6 (64) with 8 cores.

Typical for communication: trying to distribute the job to 8 cores, doing the calculations,
getting the data back, assemble it -> longer then the calculations on one core.

Cheers,

Rainer

> 
> Thanks in advance,
> 
> Mauricio
> 


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk/6yCEACgkQoYgNqgF2egrrmQCcDcYv1V1PXlO+oxHQ7wUbgait
bOoAn0tWYrIVxikHz3a4Ws2eHtJn2ATB
=4IMK
-----END PGP SIGNATURE-----