2
--------------------------------
3
LAMMPS Intel(R) Package
4
--------------------------------
6
W. Michael Brown (Intel)
7
michael.w.brown at intel.com
11
-----------------------------------------------------------------------------
13
This package is based on the USER-OMP package and provides LAMMPS styles that:
15
1. include support for single and mixed precision in addition to double.
16
2. include modifications to support vectorization for key routines
17
3. include modifications to support offload to Intel(R) Xeon Phi(TM)
20
-----------------------------------------------------------------------------
22
When using the suffix command with "intel", intel styles will be used if they
23
exist; if they do not, and the USER-OMP package is installed and an omp version
24
exists, that style will be used. For example, in the case the USER-OMP package
27
kspace_style pppm/intel 1e-4
31
kspace_style pppm/omp 1e-4
33
because no pppm style has been implemented for the Intel package.
35
-----------------------------------------------------------------------------
37
In order to use offload to Intel(R) Xeon Phi(TM) coprocessors, the flag
38
-DLMP_INTEL_OFFLOAD should be set in the Makefile. Offload requires the use of
41
-----------------------------------------------------------------------------
43
For portability reasons, vectorization directives are currently only enabled
44
for Intel compilers. Using other compilers may result in significantly
47
-----------------------------------------------------------------------------
49
By default, when running with offload to Intel(R) coprocessors, affinity
50
for host MPI tasks and OpenMP threads is set automatically within the code.
51
This currently requires the use of system calls. To disable at build time,
52
compile with -DINTEL_OFFLOAD_NOAFFINITY.