# heterodb/pg-strom

PG-Strom - Master development repository

Repository: https://github.com/heterodb/pg-strom
Canonical: https://ross.abutalabs.com/products/pg-strom
Homepage: http://heterodb.github.io/pg-strom/
Language: C
License: NOASSERTION
License Family: other
Last push: 2026-08-20T15:51:47+00:00

## Health v2 (maintenance only)
Score: 76/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 35, longevity 100
- inputs: {"age_days": 4529, "days_push": 13, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1408, forks 174 (observed 2026-08-28T04:04:38.530800+00:00)

## What it is
PG-Strom is a PostgreSQL extension that accelerates SQL analytics and batch workloads using GPU devices, NVMe-SSD storage, and Apache Arrow columnar files. It transparently offloads scans, joins, and aggregations to the GPU, including GPUDirect SQL that reads data from NVMe storage directly to the GPU.

## Use cases
- accelerate analytical SQL queries on PostgreSQL with a GPU
- run large-scale aggregations and batch processing faster
- query Apache Arrow columnar files via PostgreSQL foreign data wrapper
- read NVMe-SSD data directly into the GPU bypassing CPU and RAM
- offload WHERE-clause scans, joins, and GROUP BY to GPU cores
- speed up PostGIS spatial functions and GiST index searches on GPU

## When to choose
- you have heavy analytics or batch SQL workloads on PostgreSQL and GPU hardware available
- you want transparent GPU acceleration without rewriting applications
- you need to aggregate or search very large datasets stored as Apache Arrow files
- your bottleneck is scan/join/aggregation throughput on big data

## When to avoid
- your workload is mostly small OLTP transactions rather than analytics
- you have no NVIDIA GPU or NVMe-SSD hardware available
- you need a fully managed or cloud-native database solution
- you require frequent per-row updates on columnar Arrow data

## Facets
- artifact type: library
- maturity: active
- function: database, gpu-computing, search-engine
- domain: databases, analytics, big-data, gpu-computing
- platform: self-hosted
- tags: postgresql-extension, gpu-acceleration, sql-analytics, apache-arrow, gpudirect, nvme-ssd, columnar-storage, linux, gpu

## Member repositories
- heterodb/pg-strom (main) score 76

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:38.530800+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 2026-08-30T04:38:35.407717+00:00, confidence not recorded.
  - readme: https://github.com/heterodb/pg-strom (fetched 2026-08-28T04:04:38.530800+00:00, sha 034c774741b6)
  - homepage: http://heterodb.github.io/pg-strom/ (fetched 2026-08-29T11:52:02.933471+00:00, sha b8450f461fc2)
- Data as of 2026-08-30T08:39:29.467469+00:00.