ZLUDA is a drop-in replacement for CUDA on non-NVIDIA GPU. ZLUDA allows to run unmodified CUDA applications using non-NVIDIA GPUs with near-native performance.
ZLUDA is work in progress. Follow development here and say hi on Discord. For more details see the announcement: https://vosen.github.io/ZLUDA/blog/zludas-third-life/
Warning: ZLUDA is under heavy development (see news here). Instructions below might not work.
You should have the most recent ROCm installed.
Run your application like this:
<ZLUDA_DIRECTORY>\zluda_with.exe -- <APPLICATION> <APPLICATIONS_ARGUMENTS>
Run your application like this:
LD_LIBRARY_PATH=<ZLUDA_DIRECTORY> <APPLICATION> <APPLICATIONS_ARGUMENTS>
Warning: ZLUDA is under heavy development (see news here). Instructions below might not work.
Note: This repo has submodules. Make sure to recurse submodules when cloning this repo, e.g.: git clone --recursive https://github.com/vosen/ZLUDA.git
You should have a relatively recent version of Rust installed, then you just do:
cargo build --release
in the main directory of the project.
If you are building on Linux you must also symlink (or rename) the ZLUDA output binaries after ZLUDA build finishes:
ln -s libnvcuda.so target/release/libcuda.so
ln -s libnvcuda.so target/release/libcuda.so.1
ln -s libnvml.so target/release/libnvidia-ml.so
If you want to develop ZLUDA itself, read CONTRIBUTING.md, it contains instructions how to set up dependencies and run tests
This software is dual-licensed under either the Apache 2.0 license or the MIT license. See LICENSE-APACHE or LICENSE-MIT for details