[gmx-developers] The equestion on the implement of thread parallel in gromacs 4.5+

[Szilárd Páll]

Hi,

There are two multithreading schemes in the upcoming v4.6:
- thread_mpi - multi-threaded MPI implementation, drop-in replacement
for MPI (since 4.5)
- OpenMP (new in 4.6)

For (some) more details see:
http://www.gromacs.org/Documentation/Acceleration_and_parallelization?highlight=acceleration

Cheers,
--
Szilárd


On Tue, May 1, 2012 at 4:34 PM, Yorquant Wang <wangykoo at gmail.com> wrote:
> Hi all:
>     I recently want to add some myself pull code in the gromacs4.5.4 kernel.
> When I read the gmx's scoure code I am confused by how GMX4.5 realize its
> parallel. I know GMX4.5 can do   multi-thread parallel computation in one
> nodes without using external MPI. But  I don't know if gmx4.5 still has to
> use some inner MPI libraries, for example some thread-MPI library which
> provides a multithreaded MPI implementation or it just gets access to shared
> memory directly  for data exchanging. Is there anyone who can give me some
> instruction? Thank you so much.
>
> best wishes
>
> Yorquant .Wang
> ShangHai Jiaotong University
> ShangHai China
>
>
> --
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.