# google-coral/coralnpu

A machine learning accelerator core designed for energy-efficient AI at the edge.

Repository: https://github.com/google-coral/coralnpu
Canonical: https://ross.abutalabs.com/products/coralnpu
Language: Emacs Lisp
License: Apache-2.0
License Family: permissive
Last push: 2026-08-26T20:54:35+00:00

## Health v2 (maintenance only)
Score: 73/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 69, longevity 23
- inputs: {"age_days": 335, "days_push": 7, "days_rel": 128, "gap_med": 89.0, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2526, forks 325 (observed 2026-08-28T04:06:58.582520+00:00)

## What it is
Coral NPU is an open-source neural processing unit (NPU) hardware IP core from Google Research, built on the 32-bit RISC-V ISA with matrix, vector, and scalar processors for energy-efficient ML inference. It is designed for integration into ultra-low-power SoCs targeting wearables like hearables, AR glasses, and smartwatches, and includes RTL, simulators, and test suites.

## Use cases
- integrate an NPU into a low-power SoC for wearables
- run on-device ML inference on AR glasses or smartwatches
- simulate RISC-V based ML accelerator RTL with Verilator
- verify accelerator design with cocotb and UVM testbenches
- build energy-efficient edge AI hardware
- compile and run ML kernels on a RISC-V vector processor

## When to choose
- you are designing an ultra-low-power SoC and need licensable open-source NPU IP
- you want a RISC-V based ML accelerator you can customize and verify yourself
- you target wearables or edge devices where energy efficiency is critical

## When to avoid
- you need a ready-made USB accelerator or finished consumer product rather than IP to integrate
- you lack hardware/RTL design resources for SoC integration and verification
- you need large-scale cloud or datacenter ML acceleration

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, llm-inference, embedded, simulation, testing
- domain: machine-learning, embedded-systems, hardware, gpu-computing
- platform: embedded, iot, cpp, python
- tags: npu, risc-v, rtl, hardware-ip, edge-ai, verilator, soc-integration, wearables, open-source-hardware

## Member repositories
- google-coral/coralnpu (main) score 73

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:58.582520+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 2026-08-30T02:25:28.949353+00:00, confidence not recorded.
  - readme: https://github.com/google-coral/coralnpu (fetched 2026-08-28T04:06:58.582520+00:00, sha 92e907c5ddde)
- Data as of 2026-08-30T08:39:29.467469+00:00.