[petsc-users] [EXTERNAL] Re: Using MAGMA subroutines instead of LAPACK in SLEPc

[Harshad Sahasrabudhe]

Hi,

Yes, MAGMA's API is not exactly the same. There are a few additional 
arguments in the functions related to number of GPU etc. So it cannot be 
directly linked instead of LAPACK.

Hacking SLEPc/PETSc code seems to be the easier way to go. But I can't 
find the exact place where the LAPACK functions are called.

Thanks,
Harshad

On 06/20/2013 04:49 PM, Teranishi, Keita wrote:
> Hi,
>
> As far as I know, Magma's API us not exactly the same as the original 
> LAPACK.  You may need to hack either Magma's Fortran API or internal 
> SLEPC/PETSc source.
>
> Thanks,
> ---------------------------------------------
> Keita Teranishi
> R&D Principal Staff Member
> Sandia National Laboratories
> Livermore, CA 94551
>
> From: Matthew Knepley <knepley at gmail.com <mailto:knepley at gmail.com>>
> Date: Thursday, June 20, 2013 1:42 PM
> To: Harshad Sahasrabudhe <hsahasra at purdue.edu 
> <mailto:hsahasra at purdue.edu>>
> Cc: "petsc-users at mcs.anl.gov <mailto:petsc-users at mcs.anl.gov>" 
> <petsc-users at mcs.anl.gov <mailto:petsc-users at mcs.anl.gov>>
> Subject: [EXTERNAL] Re: [petsc-users] Using MAGMA subroutines instead 
> of LAPACK in SLEPc
>
> On Thu, Jun 20, 2013 at 10:24 PM, Harshad Sahasrabudhe 
> <hsahasra at purdue.edu <mailto:hsahasra at purdue.edu>> wrote:
>
>     Hi,
>
>     I'm trying to run the SLEPc Eigensolver on a machine with CPUs &
>     GPUs. I found that a package called MAGMA implements LAPACK
>     functions on heterogeneous systems (CPU + GPU). MAGMA has
>     functions very similar to LAPACK so that we can go and replace a
>     call to a LAPACK function with a call to a MAGMA function.
>
>     Now my question is, where can I apply such a hack inside
>     SLEPc/PETSc so that the Eigensolver in SLEPc actually uses MAGMA
>     functions instead of LAPACK functions? I'm using the PETSc 3.3-p6
>     and SLEPc 3.3-p3 versions.
>
>     Any help would be greatly appreciated!
>
>
> You need to do this in PETSc configure. Use 
> --with-f-blas-lapack-lib=[<MAGMA libs>]. We have never tried this, so 
> I don't
> know if they obey the entire interface.
>
>   Thanks
>
>     Thanks and Regards,
>     Harshad
>
>
>
>
> -- 
> What most experimenters take for granted before they begin their 
> experiments is infinitely more interesting than any results to which 
> their experiments lead.
> -- Norbert Wiener

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20130620/81c89478/attachment-0001.html>