>
>  Is your plan to use a single MPI process (that is the same as a
> sequential program) and ONE GPU for the solver or do you wish to have MPI
> parallelism for the linear solver where each MPI process uses a GPU for its
> part?


For now, I want to serialize the matrices and vectors and offload them to 1
GPU from the root process. Then distribute the result later.

Harshad

On Sat, May 30, 2015 at 10:21 PM, Barry Smith <[email protected]> wrote:

>
> > On May 30, 2015, at 8:44 PM, Harshad Sahasrabudhe <[email protected]>
> wrote:
> >
> > Hi,
> >
> > I need to offload the linear solves in a time stepper to GPUs. I'll be
> using the MAGMA sparse iterative solvers. What is the simplest way to
> achieve this? Is is better to write a new KSPType or is there a KSP wrapper
> which I can use?
>
>    We do not have any KSPType for the MAGMA sparse iterative solvers.
>
>    Is your plan to use a single MPI process (that is the same as a
> sequential program) and ONE GPU for the solver or do you wish to have MPI
> parallelism for the linear solver where each MPI process uses a GPU for its
> part?
>
>   Barry
>
> >
> > Many thanks,
> > Harshad
>
>

Reply via email to