inducer/pyopencl

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1149 stars · 248 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

PyOpenCL is a Python wrapper providing Pythonic access to the OpenCL parallel computation API, letting you run kernels on GPUs and other massively parallel devices. It includes multi-dimensional GPU arrays, parallel algorithms like reductions and prefix sums, and a fast C++ base layer with automatic error checking.

Use cases

• run compute kernels on the GPU from Python
• accelerate numpy-like array operations with OpenCL
• write parallel reductions and prefix sums on GPU
• use OpenCL from Python with automatic error handling
• do scientific computing on heterogeneous hardware
• interoperate OpenCL with OpenGL from Python

When to choose

• you need vendor-neutral GPU computing across Nvidia, AMD, Apple, and Intel
• you want full access to the OpenCL API from Python
• you need GPU arrays and parallel algorithms with a numpy-friendly workflow

When to avoid

• you only target Nvidia GPUs and prefer CUDA/PyCUDA
• you want a higher-level deep learning framework rather than raw GPU compute
• your environment has no OpenCL implementation available

Facets

library · maturity stable

gpu-computing sdk math concurrency gpu-computing developer-tools performance python windows cross-platform opencl pyopencl parallel-computing heterogeneous-computing numpy gpgpu scientific-computing arrays cuda opengl-interop linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="inducer/pyopencl")