JuliaGPU/OpenCL.jl

OpenCL Julia bindings

What it solves

OpenCL.jl provides a high-level Julia interface for the OpenCL parallel computation API, reducing the effort required to program hardware accelerators like GPUs, FPGAs, DSPs, and multicore CPUs.

How it works

The package allows users to write kernels in two ways: traditionally via OpenCL C source code, or natively using Julia functions as kernels if the platform supports SPIR-V. It manages data transfer between the host and device memory using CLArray and provides tools to build and execute kernels on available hardware platforms.

Who it’s for

Developers and researchers who need to execute parallel computations on diverse hardware accelerators using the Julia language.

Highlights

  • Multi-platform support: Works across GPUs, FPGAs, DSPs, and multicore CPUs.
  • Native Julia kernels: Support for using Julia functions as kernels via SPIR-V.
  • CLArray: Simplified data management for hardware accelerators.
  • Complete solution: Aimed at providing a complete OpenCL programming experience similar to PyOpenCL.

Related

  • Project
    JuliaGPU/CUDA.jl

    A programming interface for NVIDIA CUDA GPUs in Julia, featuring a compiler for Julia-native kernels and wrappers for various CUDA libraries.

  • Project
    omlins/ParallelStencil.jl

    A Julia library for architecture-agnostic, high-performance parallel stencil computations that allows domain scientists to write high-level code deployable on both CPUs and GPUs.

  • Project
    JuliaGPU/AMDGPU.jl

    A Julia package that enables programming and executing computations on AMD GPUs using the ROCm platform.

  • Project
    KhronosGroup/OpenCL-Headers

    The official C language headers for the OpenCL API, providing a standardized interface for developing high-performance applications on heterogeneous computing platforms.

  • Project
    JuliaGPU/Metal.jl

    A Julia package that enables GPU programming on macOS using the Metal framework, specifically for M-series Apple Silicon chips.