Environmental Modules
Environmental Modules is only available on the CentOS 7 environment of the NIG supercomputer. It is not available on the Ubuntu Linux environment. |
The Environment Modules package is a tool that lets users switch the version of the C/C++ compiler, etc. used with the user permission.
In the NIG supercomputer, we use it included in BCM and is pre-installed in our system.
Reference
Listing Available Modules
List of available packages by using the module avail command.
$ module avail
------------------------------------------------------------------------- /cm/local/modulefiles -------------------------------------------------------------------------
cluster-tools/8.2 dot gcc/9.2.0 java/1.6.0_45 java/11.0.3 module-info openmpi/mlnx/gcc/64/4.0.3rc4 r/3.5.2 ruby/2.6.5
cmd freeipmi/1.6.2 git/2.26.2 java/1.8.0_45 lua/5.3.5 null python/3.7.2 ruby/1.9.3 shared
docker/18.09.8 gcc/8.2.0 ipmitool/1.8.18 java/1.8.0_202 module-git openldap python2 ruby/2.6.0 singularity/3.4.1
------------------------------------------------------------------------ /cm/shared/modulefiles -------------------------------------------------------------------------
blacs/openmpi/gcc/64/1.1patch03 cuda91/profiler/9.1.85 intel/compiler/64/16.0.4/2016.4.258 keras-py27-cuda10.1-gcc/2.3.1
blas/gcc/64/3.8.0 cuda91/toolkit/9.1.85 intel/compiler/64/2017/17.0.8 keras-py36-cuda10.1-gcc/2.3.1
bonnie++/1.97.3 cuda92/blas/9.2.88 intel/compiler/64/2018/18.0.5 keras/2.3.1
chainer-py27-cuda10.1-gcc/6.5.0 cuda92/fft/9.2.88 intel/daal/64/2016.4/2016.4.258 lapack/gcc/64/3.8.0
chainer-py36-cuda10.1-gcc/7.0.0 cuda92/nsight/9.2.88 intel/daal/64/2017/8.262 ml-pythondeps-py27-cuda10.1-gcc/3.0.0
chainer/6.5.0 cuda92/profiler/9.2.88 intel/daal/64/2018/4.274 ml-pythondeps-py36-cuda10.1-gcc/3.0.0
cm-eigen3/3.3.7 cuda92/toolkit/9.2.88 intel/gdb/64/7.8.0/2016.4.258 mpich/ge/gcc/64/3.3
cm-ml-python3deps/3.0.0 cudnn/7.0 intel/gdb/64/2017/8.262 mvapich2/gcc/64/2.3
cm-ml-pythondeps/3.0.0 cudnn/7.5.1 intel/gdb/64/2018/5.274 mxnet-py27-cuda10.1-gcc/1.5.1
cm-pmix3/3.1.4 cudnn7.6-cuda10.1/7.6.5.32 intel/ipp/64/9.0.4/2016.4.258 mxnet-py36-cuda10.1-gcc/1.5.1
cub-cuda10.1/1.8.0 default-environment intel/ipp/64/2017/8.262 mxnet/1.5.1
cub/1.8.0 dynet-py27-cuda10.1-gcc/2.1 intel/ipp/64/2018/4.274 nccl/1.3.4
cuda10.0/blas/10.0.130 dynet-py36-cuda10.1-gcc/2.1 intel/itac/2017/8.034 nccl2-cuda10.1-gcc/2.5.6
cuda10.0/fft/10.0.130 fastai-py36-cuda10.1-gcc/1.0.60 intel/itac/2018/4.025 nccl2/2.5.6
cuda10.0/nsight/10.0.130 fftw2/openmpi/gcc/64/double/2.1.5 intel/mkl/64(default) netcdf/gcc/64/4.6.1
cuda10.0/profiler/10.0.130 fftw2/openmpi/gcc/64/float/2.1.5 intel/mkl/64/11.3.4/2016.4.258 netperf/2.7.0
cuda10.0/toolkit/10.0.130 fftw3/openmpi/gcc/64/3.3.8 intel/mkl/64/2017/8.262 openblas/dynamic(default)
cuda10.1/blas/10.1.243 gcc5/5.5.0 intel/mkl/64/2018/4.274 openblas/dynamic/0.2.20
cuda10.1/fft/10.1.243 gcc8/8.2.0 intel/mkl/mic/11.3.4/2016.4.258 opencv3-py27-cuda10.1-gcc/3.4.8
cuda10.1/nsight/10.1.243 gdb/8.2 intel/mkl/mic/2017/8.262 opencv3-py36-cuda10.1-gcc/3.4.8
cuda10.1/profiler/10.1.243 globalarrays/openmpi/gcc/64/5.7 intel/mpi/32/2017/8.262 openmpi/cuda/64/3.1.4
cuda10.1/toolkit/10.1.243 hdf5/1.10.1 intel/mpi/32/2018/4.274 openmpi/gcc/64/1.10.7
cuda11.2/blas/11.2.0 hdf5_18/1.8.20 intel/mpi/64(default) pgi/64/19.4
cuda11.2/fft/11.2.0 horovod-mxnet-py27-cuda10.1-gcc/0.18.2 intel/mpi/64/5.1.3/2016.4.258 protobuf/3.7.1
cuda11.2/toolkit/11.2.0 horovod-mxnet-py36-cuda10.1-gcc/0.18.2 intel/mpi/64/2017/8.262 protobuf3-gcc/3.7.1
cuda80/blas/8.0.61 horovod-pytorch-py27-cuda10.1-gcc/0.18.2 intel/mpi/64/2018/4.274 protobuf3-gcc8/3.7.1
cuda80/fft/8.0.61 horovod-pytorch-py36-cuda10.1-gcc/0.18.2 intel/mpi/mic/5.1.3/2016.4.258 pytorch-py27-cuda10.1-gcc/1.3.1
cuda80/nsight/8.0.61 horovod-tensorflow-py27-cuda10.1-gcc/0.18.2 intel/mpi/mic/2017/8.262 pytorch-py36-cuda10.1-gcc/1.3.1
cuda80/profiler/8.0.61 horovod-tensorflow-py36-cuda10.1-gcc/0.18.2 intel/tbb/32/4.4.6/2016.4.258 pytorch/1.3.1
cuda80/toolkit/8.0.61 horovod/0.18.2 intel/tbb/32/2017/8.262 scalapack/openmpi/gcc/64/2.0.2
cuda90/blas/9.0.176 hpcx/2.4.0 intel/tbb/32/2018/4.274 tensorflow-py27-cuda10.1-gcc/1.14.0
cuda90/fft/9.0.176 hpl/2.2 intel/tbb/64(default) tensorflow-py36-cuda10.1-gcc/1.14.0
cuda90/nsight/9.0.176 hwloc/1.11.11 intel/tbb/64/4.4.6/2016.4.258 tensorflow2-py36-cuda10.1-gcc/2.0.0
cuda90/profiler/9.0.176 intel-tbb-oss/ia32/2019_20191006oss intel/tbb/64/2017/8.262 tensorrt-cuda10.1-gcc/6.0.1.5
cuda90/toolkit/9.0.176 intel-tbb-oss/intel64/2019_20191006oss intel/tbb/64/2018/4.274 theano-py27-cuda10.1-gcc/1.0.4
cuda91/blas/9.1.85 intel/compiler/32/2017/17.0.8 intel/tbb/mic/4.4.6/2016.4.258 theano-py36-cuda10.1-gcc/1.0.4
cuda91/fft/9.1.85 intel/compiler/32/2018/18.0.5 intel/tbb/mic/2017/8.262 theano/1.0.4
cuda91/nsight/9.1.85 intel/compiler/64(default) iozone/3_482 xgboost-py36-cuda10.1-gcc/0.90
$
Using Modules
module load
By default, the gcc version 4 series included in CentOS 7 is available.
$ gcc --version
gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-28)
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
When you use the module load command, for example, it is possible to switch to gcc version 8 series.
$ module load gcc8
$ gcc --version
gcc (GCC) 8.2.0
Copyright (C) 2018 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
module unload
The module unload command can be used to restore the original state.
$ module unload gcc8
$ gcc --version
gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-28)
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
$