TheHPXProject/hpx

Health v2 · maintenance only

Adoption not part of the score

2885 stars · 550 forks observed · 2026-08-28

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

HPX is a C++ Standard Library for parallelism and concurrency that implements C++ standard parallel facilities and extends them to distributed systems. It is the first runtime implementation of the ParalleX execution model, enabling fine-grained asynchronous programming with futures and dataflow at scales from laptops to exascale clusters.

Use cases

• write fully asynchronous C++ code with futures and dataflow
• parallelize C++ applications across many cores
• build distributed HPC applications on clusters
• migrate from OpenMP, TBB, or MPI to a unified API
• scale fine-grained parallel workloads to millions of threads
• run the same code locally and across distributed nodes

When to choose

• you need a standards-oriented C++ API for parallel and distributed computing
• you want fine-grained, latency-hiding task parallelism beyond what OpenMP offers
• you are targeting large-scale HPC systems or clusters
• you want unified syntax for local and remote operations

When to avoid

• you need a simple threading library with minimal dependencies
• your project is not C++ or cannot tolerate long compile times
• you only need basic multithreading already covered by the C++ standard library
• you need a lightweight runtime for embedded or resource-constrained devices

Facets

library · maturity stable

concurrency sdk developer-tools microservices performance gpu-computing developer-tools cpp windows cross-platform hpc parallelism parallex futures task-runtime distributed-computing async cpp-standard-library linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="TheHPXProject/hpx")