Skip to content

Latest commit

 

History

History
16 lines (12 loc) · 691 Bytes

README.md

File metadata and controls

16 lines (12 loc) · 691 Bytes

StandardML-GPU

Adding fast CUDA bindings to Standard ML

For more information, check out our website at https://rohany.github.io/StandardML-GPU

To see SML-GPU in action, clone the repository with git clone https://github.com/rohany/StandardML-GPU/.

Install the dependencies

sudo apt-get install mlton

Follow the Nvidia CUDA setup guide to install CUDA. Then head over to the tests directory, and make sure that cuda path in the Makefiles is the same as the path to your installation. Also ensure that your CUDA compute compatability in the lib makefile is the highest that that of your card supports.

Then just run make, ./<executable> <input size>, and see the speedups!