Cixelyn edited this page Sep 13, 2010 · 12 revisions
Clone this wiki locally

These set of functions present several methods for calculating the solution to the equation AX=B on both the CPU and the GPU using the conjugate gradient method

This project was conducted as a part of the 6.936 MIT Class
You can also read our research paper over at Google Docs detailing all the work and algorithms.

Included in the source tree are the code used for cranking computations, as well as some auxiliary python code for test case generation and also a test data file from our experimental trials.

Feel free to take the code and integrate it into your own projects in order to accelerate your computations. Large computations should see close to a 20x speedup! Make sure to read the readme for help in using the included code, or look through the wiki page on function calls. If you get stuck, need help, or just want to chat, you’re welcome to email the authors of the paper.