rust-gpu-tools

An abstraction library to run kernels on both CUDA and OpenCL.

Example

You need to write the code that interacts with the GPU only once. Below is such code that runs a kernel on CUDA and/or OpenCL. For a full working example, please see the examples directory. You can run it via cargo run --example add.

let closures = program_closures!(|program, _args| -> Result<Vec<u32>, GPUError> {
    // Make sure the input data has the same length.
    assert_eq!(aa.len(), bb.len());
    let length = aa.len();

    // Copy the data to the GPU.
    let aa_buffer = program.create_buffer_from_slice(&aa)?;
    let bb_buffer = program.create_buffer_from_slice(&bb)?;

    // The result buffer has the same length as the input buffers.
    let result_buffer = unsafe { program.create_buffer::<u32>(length)? };

    // Get the kernel.
    let kernel = program.create_kernel("add", 8, 4)?;

    // Execute the kernel.
    kernel
        .arg(&(length as u32))
        .arg(&aa_buffer)
        .arg(&bb_buffer)
        .arg(&result_buffer)
        .run()?;

    // Get the resulting data.
    let mut result = vec![0u32; length];
    program.read_into_buffer(&result_buffer, &mut result)?;

    Ok(result)
});

Environment variables

• RUST_GPU_TOOLS_CUSTOM_GPU

  rust-gpu-tools has a hard-coded list of GPUs and their CUDA core count. If your card is not part of that list, you can add it via RUST_GPU_TOOLS_CUSTOM_GPU. The value is a comma separated list of name:cores. Example:

   RUST_GPU_TOOLS_CUSTOM_GPU="GeForce RTX 2080 Ti:4352,GeForce GTX 1060:1280"
  

License

Licensed under either of

• Apache License, Version 2.0, (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
• MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)

at your option.

Contribution

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.