https://github.com/larsgeb/m1-gpu-cpp

Metal Shading Language on Apple M1's GPU for scientific C++.
https://github.com/larsgeb/m1-gpu-cpp

clang cpp cpp17 gpu-acceleration gpu-computing m1-mac metal metal-cpp objective-c scientific-computing

Last synced: 3 months ago
JSON representation

Metal Shading Language on Apple M1's GPU for scientific C++.

• Host: GitHub
• URL: https://github.com/larsgeb/m1-gpu-cpp
• Owner: larsgeb
• License: other
• Created: 2022-04-20T12:20:59.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2023-10-20T19:36:42.000Z (over 1 year ago)
• Last Synced: 2024-10-09T08:08:15.091Z (4 months ago)
• Topics: clang, cpp, cpp17, gpu-acceleration, gpu-computing, m1-mac, metal, metal-cpp, objective-c, scientific-computing
• Language: C++
• Homepage:
• Size: 10.9 MB
• Stars: 79
• Watchers: 3
• Forks: 15
• Open Issues: 4
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# M1 GPUs for (scientific) computations in C++

In this repo, I explore the capabilities of the Metal Shading Language 

combined with C++, specifically for use of accelerating scientific codes.

This repo accompaines "Seamless GPU acceleration for C++ based physics using the M1's

unified processing units, a case study for elastic wave propagation and full-waveform

inversion". If you use VSCode, all instructions to compile can be found in `.vscode`.

I also wrote a few blog posts about the work in this repo:

- [Getting started](https://larsgeb.github.io/2022/04/20/m1-gpu.html)

- [SAXPY and FD](https://larsgeb.github.io/2022/04/22/m1-gpu.html)

However, if you just want to try out the code, make sure you install `llvm` 

and `libomp` using brew and Xcode and its command line tools. By cloning the 

repo and opening it in VSCode you should have all build configurations.

Additionally, I wrote a preprint (with the actual paper currently in peer-review) on how

to use MSL for modelling of PDEs [available on arXiv](https://arxiv.org/abs/2206.01791).