Skip to content
/ vuda Public

VUDA is a header-only library based on Vulkan that provides a CUDA Runtime API interface for writing GPU-accelerated applications.

License

Notifications You must be signed in to change notification settings

jgbit/vuda

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 

Repository files navigation

VUDA

VUDA is a header-only library based on Vulkan that provides a CUDA Runtime API interface for writing GPU-accelerated applications.

Documentation

VUDA is based on the Vulkan API. The functionality of VUDA conforms (as much as possible) to the specification of the CUDA runtime. For normal usage consult the reference guide for the NVIDIA CUDA Runtime API, otherwise check the VUDA wiki:

Usage

All VUDA functionality can be accessed by including vuda.hpp and using its namespace vuda::. Alternatively, one can utilize vuda_runtime.hpp which wraps and redirect all CUDA functionality.

#if defined(__NVCC__)
    #include <cuda_runtime.h>
#else
    #include <vuda_runtime.hpp>
#endif

int main(void)
{
    // assign a device to the thread
    cudaSetDevice(0);
    // allocate memory on the device
    const int N = 5000;
    int a[N], b[N], c[N];
    for(int i = 0; i < N; ++i)
    {
        a[i] = -i;
        b[i] = i * i;
    }
    int *dev_a, *dev_b, *dev_c;
    cudaMalloc((void**)&dev_a, N * sizeof(int));
    cudaMalloc((void**)&dev_b, N * sizeof(int));
    cudaMalloc((void**)&dev_c, N * sizeof(int));
    // copy the arrays a and b to the device
    cudaMemcpy(dev_a, a, N * sizeof(int), cudaMemcpyHostToDevice);
    cudaMemcpy(dev_b, b, N * sizeof(int), cudaMemcpyHostToDevice);
    // run kernel (vulkan shader module)
    const int blocks = 128;
    const int threads = 128;
#if defined(__NVCC__)
    add<<<blocks, threads>>>(dev_a, dev_b, dev_c, N);
#else
    const int stream_id = 0;
    vuda::launchKernel("add.spv", "main", stream_id, blocks, threads, dev_a, dev_b, dev_c, N);
#endif
    // copy result to host
    cudaMemcpy(c, dev_c, N * sizeof(int), cudaMemcpyDeviceToHost);

    // do something useful with the result in array c ...        

    // free memory on device
    cudaFree(dev_a);
    cudaFree(dev_b);
    cudaFree(dev_c);
}

About

VUDA is a header-only library based on Vulkan that provides a CUDA Runtime API interface for writing GPU-accelerated applications.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages