Edits to accelerate_kokkos.txt

This commit is contained in:
Stan Moore 2017-12-18 14:50:33 -07:00
parent 8a93f63de9
commit 798d68c607
1 changed files with 11 additions and 9 deletions

View File

@ -53,9 +53,9 @@ See the [Advanced Kokkos Options] section below for a listing of all KOKKOS_ARCH
[Compile for CPU-only (MPI only, no threading):]
use a C++11 compatible compiler and
set KOKKOS_ARCH variable in /src/MAKE/OPTIONS/Makefile.kokkos_mpi_only
as described above
use a C++11 compatible compiler and set KOKKOS_ARCH variable in
/src/MAKE/OPTIONS/Makefile.kokkos_mpi_only as described above. Then do the
following:
cd lammps/src
make yes-kokkos
@ -67,8 +67,9 @@ NOTE: To build with Kokkos support for OpenMP threading, your compiler must supp
OpenMP interface. You should have one or more multi-core CPUs so that
multiple threads can be launched by each MPI task running on a CPU.
use a C++11 compatible compiler and
set KOKKOS_ARCH variable in /src/MAKE/OPTIONS/Makefile.kokkos_omp
use a C++11 compatible compiler and set KOKKOS_ARCH variable in
/src/MAKE/OPTIONS/Makefile.kokkos_omp as described above. Then do the
following:
cd lammps/src
make yes-kokkos
@ -76,7 +77,7 @@ make kokkos_omp :pre
[Compile for Intel KNL Xeon Phi (Intel Compiler, OpenMPI):]
use a C++11 compatible compiler and
use a C++11 compatible compiler and do the following:
cd lammps/src
make yes-kokkos
@ -89,8 +90,9 @@ version 7.5 or later must be installed on your system. See the
discussion for the "GPU"_accelerate_gpu.html package for details of
how to check and do this.
use a C++11 compatible compiler and
set KOKKOS_ARCH variable in /src/MAKE/OPTIONS/Makefile.kokkos_cuda_mpi for both GPU and CPU
use a C++11 compatible compiler and set KOKKOS_ARCH variable in
/src/MAKE/OPTIONS/Makefile.kokkos_cuda_mpi for both GPU and CPU as described
above. Then do the following:
cd lammps/src
make yes-kokkos
@ -453,4 +455,4 @@ with the {enable_lambda} option when using GPUs.
Currently, there are no precision options with the KOKKOS
package. All compilation and computation is performed in double
precision.
precision.