User Tools

Site Tools


cluster:73

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
cluster:73 [2021/03/16 19:00]
hmeij07 [OpenHPC]
cluster:73 [2021/04/16 19:57]
hmeij07 [LAMMPS CPU/GPU]
Line 26: Line 26:
   * http://repos.openhpc.community/OpenHPC/2.0RC1/CentOS_8/x86_64/   * http://repos.openhpc.community/OpenHPC/2.0RC1/CentOS_8/x86_64/
   * http://repos.openhpc.community/ohpc-1.3/1.3.9/base/CentOS_7/x86_64/   * http://repos.openhpc.community/ohpc-1.3/1.3.9/base/CentOS_7/x86_64/
 +  * https://github.com/warewulf/warewulf3 (master git zip file)
  
 ====== OneAPI ====== ====== OneAPI ======
Line 1335: Line 1336:
  
 ====== LAMMPS CPU/GPU ====== ====== LAMMPS CPU/GPU ======
 +
 +** USER_QUIP **
 +
 +This is a version of lammps (29Oct20, last stable) that contains all packages with user-quip.
 +
 +It does compile with MPI (mpif90), not ifort, in CentOS7 ... was tricky, solution for MPI was the ''mpifort --showme:link'' answer
 +
 +  * Requires CentOS 7 so queues: mwgpu, amber128, exx96 for cpu-only binaries
 +  * Rwequires CentOS7 plus Cuda 10.2 so queue: exx96 for cpu+gpu binaries
 +
 +Read, if you wish  
 +
 +  * /share/apps/CENTOS7/lammps/QUIP-public.install 
 +  * /share/apps/CENTOS7/lammps/29Oct2020.install 
 +
 +**Environments**
 +
 +<code>
 +
 +# serial - cpu only (although quippy compiled, it refused to install)
 +# MPI STUBS is compiled in
 +
 +export PATH=/share/apps/CENTOS7/python/3.8.3/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/python/3.8.3/lib:$LD_LIBRARY_PATH
 +
 +/share/apps/CENTOS7/lammps/29Oct2020/lmp_serial-with-quip.cpu -h
 +
 +# mpi - cpu only
 +# gfortran_openmpi
 +
 +export QUIP_LIBRARY=/share/apps/CENTOS7/lammps/QUIP-public/9Apr2021/linux_x86_64_gfortran_openmpi
 +export PATH=/share/apps/CENTOS7/python/3.8.3/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/python/3.8.3/lib:$LD_LIBRARY_PATH
 +export PATH=/share/apps/CENTOS7/openmpi/4.0.4/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/openmpi/4.0.4/lib:$LD_LIBRARY_PATH
 +
 +/share/apps/CENTOS7/lammps/29Oct2020/lmp_mpi-with-quip.cpu -h
 +
 +
 +# mpi - cpu+gpu
 +# gfortran_openmpi arch
 +
 +export PATH=/share/apps/CENTOS7/python/3.8.3/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/python/3.8.3/lib:$LD_LIBRARY_PATH
 +export QUIP_LIBRARY=/share/apps/CENTOS7/lammps/QUIP-public/9Apr2021/linux_x86_64_gfortran_openmpi
 +export PATH=/share/apps/CENTOS7/python/3.8.3/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/python/3.8.3/lib:$LD_LIBRARY_PATH
 +export PATH=/share/apps/CENTOS7/openmpi/4.0.4/bin:$PATH
 +export LD_LIBRARY_PATH=/share/apps/CENTOS7/openmpi/4.0.4/lib:$LD_LIBRARY_PATH
 +export CUDA_HOME=/usr/local/cuda
 +export PATH=/usr/local/cuda/bin:$PATH
 +export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
 +export LD_LIBRARY_PATH=/usr/local/cuda/lib:$LD_LIBRARY_PATH
 +
 +# the "make" versions of lmp with quip do not work, deleted them
 +# did find on nvdia dev forum reference problem disappears when using cmake (vs make)
 +# these lmp binaries do work
 +/share/apps/CENTOS7/lammps/29Oct2020/single_single/bin/lmp
 +/share/apps/CENTOS7/lammps/29Oct2020/single_double/bin/lmp
 +/share/apps/CENTOS7/lammps/29Oct2020/double_double/bin/lmp
 +
 +# however I had to load non-mpi version of libquid.a into lmp
 +# witch seems ok given parallel work is driven by -suffix gpu
 +
 +# sample script: ~hmeij/k20redo/run.rtx.lammps2
 +
 +</code>
  
 ** CPU only ** -Centos6 ** CPU only ** -Centos6
cluster/73.txt ยท Last modified: 2024/05/15 12:58 by hmeij07