# uxlfoundation/oneTBB

oneAPI Threading Building Blocks (oneTBB)

Repository: https://github.com/uxlfoundation/oneTBB
Canonical: https://ross.abutalabs.com/products/onetbb
Homepage: https://uxlfoundation.github.io/oneTBB/
Language: C++
License: Apache-2.0
License Family: permissive
Topics: tbbmalloc, threading, parallelism, composability, taskparallelism, flowgraph, oneapi, tbb
Last push: 2026-08-26T16:15:05+00:00

## Health v2 (maintenance only)
Score: 89/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 69, longevity 100
- inputs: {"age_days": 3647, "days_push": 7, "days_rel": 51, "gap_med": 120, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6723, forks 1179 (observed 2026-08-28T04:09:48.087467+00:00)

## What it is
oneTBB (oneAPI Threading Building Blocks) is a C++ library for adding task-based parallelism to applications, letting developers express logical parallelism rather than manage threads directly. It provides parallel algorithms, flow graphs, concurrent containers, and a scalable memory allocator, and is an implementation of the oneAPI specification under the UXL Foundation.

## Use cases
- parallelize CPU-bound C++ code across multiple cores
- build data-parallel pipelines and task flow graphs
- replace raw std::thread management with composable task-based parallelism
- get a scalable multithreaded memory allocator (tbbmalloc)
- write portable parallel code for Windows, Linux, and macOS
- scale performance-sensitive algorithms to many-core processors

## When to choose
- you want high-level, composable parallelism in C++ without being a threading expert
- you need scalable data-parallel programming and work-stealing scheduling
- you want a mature, permissively licensed (Apache-2.0) library compatible with other threading packages
- you are targeting the oneAPI ecosystem

## When to avoid
- you need GPU or heterogeneous offload rather than CPU threading
- your project is not C++ or cannot link native libraries
- you only need simple async concurrency already covered by std::async or coroutines
- you need distributed-memory (multi-node) parallelism like MPI

## Facets
- artifact type: library
- maturity: stable
- function: concurrency, developer-tools
- domain: developer-tools, performance, cross-platform
- platform: windows, cpp, cross-platform
- tags: parallelism, task-parallelism, flow-graph, memory-allocator, oneapi, data-parallelism, multicore, linux, macos

## Member repositories
- uxlfoundation/oneTBB (main) score 89

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:48.087467+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 2026-08-29T17:42:54.097930+00:00, confidence not recorded.
  - readme: https://github.com/uxlfoundation/oneTBB (fetched 2026-08-28T04:09:48.087467+00:00, sha 5cf4693bf0d4)
  - homepage: https://uxlfoundation.github.io/oneTBB/ (fetched 2026-08-29T08:38:52.218565+00:00, sha 0a62b1dbed1a)
- Data as of 2026-08-30T08:39:29.467469+00:00.