# uber/fiber

Distributed Computing for AI Made Simple

Repository: https://github.com/uber/fiber
Canonical: https://ross.abutalabs.com/products/uber-fiber
Homepage: https://uber.github.io/fiber/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: python, distributed-computing, multiprocessing, sandbox, machine-learning
Archived: true
Last push: 2023-03-19T22:55:22+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2430, "days_push": 1263, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1047, forks 108 (observed 2026-08-28T04:03:22.218468+00:00)

## What it is
Fiber is a Python distributed computing library for modern computer clusters that exposes the same API as Python's standard multiprocessing library. It uses Nanomsg for fast inter-process communication and Docker/Kubernetes to run code on clusters without manual deployment.

## Use cases
- run multiprocessing-style code on a computer cluster
- parallelize large-scale scientific computations like POET
- scale AI model training across cluster CPUs and GPUs
- distribute reinforcement learning experiments
- run parallel workloads on Kubernetes without cluster expertise

## When to choose
- you already know Python multiprocessing and want cluster-level scaling
- you want to run parallel jobs on Kubernetes without deploying a separate framework
- you need dynamic CPU/memory/GPU resource allocation for parallel workloads

## When to avoid
- you need a stable, actively maintained project (APIs are experimental and releases stopped in 2023)
- you need general-purpose data pipeline processing like Spark or Dask
- you only need parallelism on a single machine (standard multiprocessing suffices)

## Facets
- artifact type: library
- maturity: maintenance
- function: concurrency, machine-learning, workflow-automation
- domain: microservices, machine-learning, developer-tools
- platform: python
- tags: multiprocessing, cluster-computing, parallel-computing, nanomsg, reinforcement-learning, linux, docker, kubernetes

## Member repositories
- uber/fiber (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:22.218468+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 2026-08-30T07:01:10.466219+00:00, confidence not recorded.
  - readme: https://github.com/uber/fiber (fetched 2026-08-28T04:03:22.218468+00:00, sha 1ddd40aaae8b)
  - homepage: https://uber.github.io/fiber/ (fetched 2026-08-29T13:02:52.701308+00:00, sha 92e67d52f64e)
  - registry_pypi: https://pypi.org/pypi/fiber/json (fetched 2026-08-29T13:02:52.710275+00:00, sha cb014fb5e188)
- Data as of 2026-08-30T08:39:29.467469+00:00.