Installation

M-Star CFD is a multi–GPU-based solver which contains pre-processing, solver, and post-processing environments. It is cross-platform and easy to integrate into other workflows. Explore this section to get familiar with the available capabilities.

Administration and Requirements

• System Requirements : Which operating systems and GPUs are needed to run M-Star CFD.

• Hardware Guide : How to select GPUs and hardware

• Platform Diagnostics : What’s going on under the hood.

• 3D device troubleshooting : Fix common issues with 3D rendering programs Pre and Post

Software Installation

• Windows : How to install on Windows. This is the most common setup.

• Linux (Single Node) : Instructions for Linux single-node installations.

• Linux (Multi Node) : Installation in a multi-node GPU environment.

• AWS : Instructions to run M-Star on Amazon Web Services.

• Azure : Instructions for a single Azure setup with multiple GPUs.

• Google Cloud : Create an Instance on Google Cloud.

• Containers : Run M-Star using Docker or Singularity.

Optional Packages

• Python : Optionally used for scripting M-Star Pre/Solve/Post activities, workflows, and integrations.

• Windows Python Pre Setup Guide : Recommended Python setup on Windows with VS Code IDE

• FFmpeg on Linux : Needed on Linux to save moves. If this package isn’t already on your machine, you can get it here.

• Remote visualization: Access a remote desktop using NICE DCV.