Binary_EN - CCSEPBVR/CS-IS-PBVR GitHub Wiki

Binary

The GUI program and the server side programs consisting CS/IS-PBVR run in Linux, Mac, and Windows environments, and provide sequential processing and parallelized binaries with OpenMP. In addition, massively parallelized binaries are available for supercomputers SGI8600 and Fugaku using MPI + OpenMP.
※ On a Mac environment, when executing the binary, you might see the message "'.app' is damaged and cannot be opened. It needs to be moved to the trash." In that case, run the following command in the directory where it was extracted.

cd v3.X.X_Mac
sudo xattr -rc *

CS

KVSML Converter

The load module of the KVSML Converter is available in a sequential version. It can be used by copying the load module of each environment to an appropriate directory in the PATH. The SGI8600 login node can use the load module for Linux. The configuration of the load module is as follows.(v.x.x.x is version.)

Models Parallelization Directory Name/Load Module Name
Linux 64bit Serial process v.x.x.x_Linux/kvsml-converter
Mac 64bit Serial process v.x.x.x_Mac/kvsml-converter
Windows 64bit Serial process v.x.x.x_Windows_DESKTOP/kvsml-converter

Filters

The load module of the filter program is available in a sequential version, an OpenMP version for thread parallel processing, and an MPI + OpenMP version for massively parallel processing. It can be used by copying the load module of each environment to an appropriate directory in the PATH. The configuration of the load module is as follows.

Models Parallelization Load Module Name
Linux 64bit Serial process pbvr_filter_linux
OpenMP pbvr_filter_linux_omp
Mac 64bit Serial process pbvr_filter_mac
OpenMP pbvr_filter_mac_omp
Windows 64bit Serial process pbvr_filter_win
OpenMP pbvr_filter_omp_win
SGI8600 ※1 MPI+OpenMP pbvr_filter_s86_mpi_omp
Fugaku MPI+OpenMP pbvr_filter_fugaku_mpi_omp

*1. Since the load module for the supercomputer is dedicated to the compute node, the load module for Linux is used for the login node and post-processing node configured by the Linux server.

Server

The server program is implemented in C++, and there is a sequential version, an OpenMP version for thread parallelism, and an MPI + OpenMP version for massively parallel processing. It can be used by copying the load module of each environment to an appropriate directory in the PATH. The configuration of the load module is as follows.

Models Parallelization Load Module Name
Linux 64bit Serial process pbvr_server_linux
OpenMP pbvr_server_linux_omp
Mac 64bit Serial process pbvr_server_mac
OpenMP pbvr_server_mac_omp
Windows 64bit Serial process pbvr_server_win
OpenMP pbvr_server_omp_win
SGI8600※1 MPI+OpenMP pbvr_server_s86_mpi_omp
Fugaku MPI+OpenMP pbvr_server_fugaku_mpi_omp

*1. Since the load module for the supercomputer is dedicated to the compute node, the load module for Linux is used for the login node and post-processing node configured by the Linux server.

IS

Daemon

The daemon program is implemented in C++, and an OpenMP version that uses parallel processing is available. It can be used by copying the load module of each environment to an appropriate directory in the PATH. The configuration of the load module is as follows.

Models Parallelization Load Module Name
SGI8600 OpenMP pbvr_daemon_s86
Fugaku OpenMP pbvr_daemon_fugaku

Particle Sampler

The particle sampler is implemented in C++, and an MPI + OpenMP version is available for massively parallel processing. It can be used by copying the load module of each environment to an appropriate directory in the PATH.

Models Parallelization Load Module Name
SGI8600 MPI+OpenMP pbvr_sampler_s86
Fugaku MPI+OpenMP pbvr_sampler_fugaku

Common

Client

The load module of the client program is parallelized by QThread. It can be used by copying the load module of each environment to an appropriate directory in the PATH. The configuration of the load module is as follows.

Models Load Module Name
Linux 64bit pbvr_client_linux
Mac 64bit pbvr_client_mac
Windows 64bit pbvr_client_win
⚠️ **GitHub.com Fallback** ⚠️