# inducer/pyopencl

OpenCL integration for Python, plus shiny features

Repository: https://github.com/inducer/pyopencl
Canonical: https://ross.abutalabs.com/products/pyopencl
Homepage: http://mathema.tician.de/software/pyopencl
Language: Python
License: NOASSERTION
License Family: other
Topics: python, gpu, opencl, heterogeneous-parallel-programming, nvidia, cuda, amd, performance, array, multidimensional-arrays, parallel-computing, shared-memory, parallel-algorithm, prefix-sum, reduction, sorting, pyopencl, opengl, scientific-computing
Last push: 2026-08-25T19:24:46+00:00

## Health v2 (maintenance only)
Score: 99/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 100
- inputs: {"age_days": 5628, "days_push": 8, "days_rel": 8, "gap_med": 10.5, "n_releases_24m": 15}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1149, forks 248 (observed 2026-08-28T04:03:46.296957+00:00)

## What it is
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
- artifact type: library
- maturity: stable
- function: gpu-computing, sdk, math, concurrency
- domain: gpu-computing, developer-tools, performance
- platform: python, windows, cross-platform
- tags: opencl, pyopencl, parallel-computing, heterogeneous-computing, numpy, gpgpu, scientific-computing, arrays, cuda, opengl-interop, linux, macos

## Member repositories
- inducer/pyopencl (main) score 99

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:46.296957+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:33:44.778488+00:00, confidence not recorded.
  - readme: https://github.com/inducer/pyopencl (fetched 2026-08-28T04:03:46.296957+00:00, sha 301cd1f2015b)
  - homepage: http://mathema.tician.de/software/pyopencl (fetched 2026-08-29T12:38:49.746884+00:00, sha a0eb8e40a371)
  - site_page: https://mathema.tician.de/aboutme (fetched 2026-08-29T12:38:49.749512+00:00, sha dc48d3674be5)
  - site_page: https://documen.tician.de/pyopencl (fetched 2026-08-29T12:38:49.751929+00:00, sha ff3ccc0a54cc)
- Data as of 2026-08-30T08:39:29.467469+00:00.