# google-research/circuit_training

Repository: https://github.com/google-research/circuit_training
Canonical: https://ross.abutalabs.com/products/circuit_training
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-02-11T21:34:53+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 67, release rhythm 35, longevity 100
- inputs: {"age_days": 1765, "days_push": 203, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1706, forks 273 (observed 2026-08-28T04:05:24.939784+00:00)

## What it is
AlphaChip (circuit_training) is Google's open-source framework for generating chip floorplans using distributed deep reinforcement learning, reproducing the Nature 2021 graph placement methodology. It is built on TF-Agents and TensorFlow 2.x, supporting distributed training across multiple GPUs and hundreds of data-collection actors.

## Use cases
- generate chip floorplans with deep reinforcement learning
- place netlists with hundreds of macros and millions of standard cells
- optimize wirelength, congestion, and density in chip layouts
- reproduce the AlphaChip Nature 2021 paper methodology
- run distributed RL training across multiple GPUs for placement
- research AI methods for chip design and EDA

## When to choose
- you need RL-based macro placement and floorplanning for chip design
- you want to reproduce or extend the AlphaChip methodology from research
- you need distributed training across many GPUs and actors for placement
- you work in EDA research exploring AI for chip layout

## When to avoid
- you need a general-purpose EDA flow or full place-and-route tool
- your netlists are not in clustered format or lack macro/stdcell preparation
- you need a lightweight CPU-only solution without TensorFlow
- you want a turnkey commercial chip design tool rather than a research framework

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, reinforcement-learning, simulation
- domain: machine-learning, hardware, gpu-computing
- platform: python
- tags: chip-design, floorplanning, alphachip, eda, distributed-training, tf-agents, reinforcement-learning, linux, gpu

## Member repositories
- google-research/circuit_training (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:24.939784+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 2026-08-30T03:37:28.979632+00:00, confidence not recorded.
  - readme: https://github.com/google-research/circuit_training (fetched 2026-08-28T04:05:24.939784+00:00, sha e8dac05270f2)
- Data as of 2026-08-30T08:39:29.467469+00:00.