Co-developed by Computer Engineering students at Cal Poly, faculty and AMD as a Capstone project, this accelerated computing teaching kit covers introductory topics about AMD GPUs and how to interface with them. All content and code was written with an AMD Vega in mind, but should apply to any AMD GPU supported by ROCm.
Laid out in the below table of contents are all the chapters in this teaching kit. To access any topic, simply click its name in the table of contents. This will take you to that topic's introduction. Within the introduction will be links to chapter material, example code, and a quiz to test your understanding. All material is contained within markdown files (besides in the cases of example code), and can be navigated via hyperlinks within the markdown files.
- Introduction to ROCm
- Memory and Data Locality
- OpenMP
- Translating CUDA to HIP
- Parallel Computing
- Neural networks
If you want to contribute or help with the development of this teaching kit, create a fork of this repo and open a feature branch. Then, create a pull request. We'd love to make this teaching kit the best it can be with your help!