NERSC Perlmutter

Installation on NERSC Perlmutter

Weak-coupling Many-Body Perturbation theory solver

Setup Build Environment

Code has been tested for the following NERSC programming environments:

PrgEnv-gnu/8.5.0
PrgEnv-gnu/8.6.0
PrgEnv-intel/8.5.0
PrgEnv-aocc/8.5.0
PrgEnv-cray/8.5.0

Load an environment by calling

module load PrgEnv-gnu/8.5.0

Load third-party libraries

module load cray-hdf5 cmake

Download and build the Many-Body Perturbation theory solver

We recommend building a specific released version — for example the latest stable v0.3.2 or the latest pre-release v1.0.0a1. See the releases page for newer tags.

git clone --branch v0.3.2 --depth 1 https://github.com/Green-Phys/green-mbpt
cmake -S green-mbpt -B green-mbpt-build                         \
 -DCMAKE_INSTALL_PREFIX=/path/to/install/directory              \
 -DCMAKE_BUILD_TYPE=Release
cmake --build green-mbpt-build -j 4
cmake --build green-mbpt-build -t test

git clone --branch v0.3.2 --depth 1 https://github.com/Green-Phys/green-mbpt
cmake -S green-mbpt -B green-mbpt-build                         \
 -DCMAKE_INSTALL_PREFIX=/path/to/install/directory              \
 -DCMAKE_BUILD_TYPE=Release                                     \
 -DCUSTOM_KERNELS="https://github.com/Green-Phys/green-gpu"
cmake --build green-mbpt-build -j 4
cmake --build green-mbpt-build -t test

Job script example

#!/bin/bash

#SBATCH -N <Number of nodes>
#SBATCH -C cpu
#SBATCH -q <QOS>
#SBATCH -J <Job name>
#SBATCH -A <Account>
#SBATCH -t 0:30:0

# OpenMP settings:
export OMP_NUM_THREADS=1
export OMP_PLACES=threads
export OMP_PROC_BIND=spread
export HDF5_USE_FILE_LOCKING=FALSE

#run the application:
srun -n 64 <build directory>/mbpt.exe --grid_file ir/1e4.h5 \
                                    --beta 100 \
                                    --scf_type GW \
                                    --itermax=1 \
                                    --results_file Si.h5 \
                                    --jobs SC

#!/bin/bash

#SBATCH -N <Number of nodes>
#SBATCH -C gpu
#SBATCH -q <QOS>
#SBATCH -J <Job name>
#SBATCH -A <Account>
#SBATCH -t 0:30:0
#SBATCH --gpus-per-task=4

# OpenMP settings:
export OMP_NUM_THREADS=1
export OMP_PLACES=threads
export OMP_PROC_BIND=spread
export HDF5_USE_FILE_LOCKING=FALSE

#run the application:
srun -n 64 <build directory>/mbpt.exe --grid_file ir/1e4.h5 \
                                    --beta 100 \
                                    --scf_type GW \
                                    --itermax=1 \
                                    --kernel GPU \
                                    --cuda_low_gpu_memory false \
                                    --cuda_low_cpu_memory false \
                                    --results_file Si.h5 \
                                    --jobs SC