boostorg/compute

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1660 stars · 339 forks observed · 2026-08-28

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

Boost.Compute is a C++ GPU/parallel-computing library built on OpenCL, providing a thin C++ wrapper over the OpenCL API plus an STL-like interface with containers, algorithms, and fancy iterators. It is header-only and lets developers run computations like sorting and reductions on GPUs and multi-core CPUs.

Use cases

• sort large arrays of floats on the GPU
• run parallel reductions and scans on OpenCL devices
• copy data between host and GPU device memory
• use STL-style algorithms like transform and accumulate on the GPU
• write portable GPGPU code in C++ without raw OpenCL boilerplate
• interoperate with the OpenCL and OpenGL APIs

When to choose

• you want an STL-like C++ interface for OpenCL GPU computing
• you need portable GPGPU code across OpenCL implementations
• you prefer a header-only library with no linking step
• you want common parallel algorithms and containers out of the box

When to avoid

• you need CUDA-specific features or NVIDIA-only optimizations
• you want actively developed cutting-edge GPU compute features
• you target platforms without OpenCL support
• you need a modern C++ GPU library like SYCL instead

Facets

library · maturity maintenance

gpu-computing math concurrency gpu-computing performance developer-tools cpp cross-platform opencl gpgpu stl-like header-only parallel-computing boost algorithms gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="boostorg/compute")