# NVIDIA/cudf

cuDF - GPU DataFrame Library

Repository: https://github.com/NVIDIA/cudf
Canonical: https://ross.abutalabs.com/products/cudf
Homepage: https://docs.nvidia.com/cudf/
Language: C++
License: Apache-2.0
License Family: permissive
Topics: gpu, rapids, cudf, arrow, cuda, pandas, dataframe, dask, data-analysis, data-science, pydata, cpp, python
Last push: 2026-08-26T23:56:20+00:00

## Health v2 (maintenance only)
Score: 95/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 87, longevity 100
- inputs: {"age_days": 3405, "days_push": 7, "days_rel": 8, "gap_med": 55, "n_releases_24m": 18}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 9734, forks 1095 (observed 2026-08-28T04:10:36.996327+00:00)

## What it is
cuDF is a GPU-accelerated DataFrame library for tabular data processing, part of NVIDIA's RAPIDS suite. It provides a pandas-compatible Python API (cudf), a zero-code-change pandas accelerator (cudf.pandas), a GPU engine for Polars, a Dask backend, and an underlying CUDA C++ core (libcudf) with Apache Arrow-compliant data structures.

## Use cases
- accelerate pandas dataframes on gpu
- process large csv datasets faster than pandas
- run pandas code unchanged on gpu with cudf.pandas
- gpu backend for polars queries
- scale dataframe processing with dask on gpus
- etl pipelines for big tabular data
- data analysis of millions of rows without running out of memory

## When to choose
- you have nvidia gpus and pandas workloads that are slow or memory-bound
- you want a drop-in pandas accelerator with zero code changes
- you need gpu-accelerated etl or dataframe operations at scale
- you want a gpu engine for polars or dask dataframes

## When to avoid
- you have no cuda-capable nvidia gpu
- your datasets are small enough that pandas is already fast enough
- you need cpu-only portability or non-nvidia accelerators
- you rely on pandas features not yet supported by the cudf api

## Facets
- artifact type: library
- maturity: stable
- function: data-science, etl, gpu-computing, database
- domain: data-science, big-data, gpu-computing, analytics
- platform: python, cpp
- tags: dataframe, pandas, rapids, cuda, arrow, dask, polars, tabular-data, data-engineering, linux, gpu, docker

## Member repositories
- NVIDIA/cudf (main) score 95

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:36.996327+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 2026-08-29T17:21:13.303984+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA/cudf (fetched 2026-08-28T04:10:36.996327+00:00, sha d30e5603619c)
  - homepage: https://docs.nvidia.com/cudf/ (fetched 2026-08-29T08:20:45.607612+00:00, sha 44136fa355b3)
- Data as of 2026-08-30T08:39:29.467469+00:00.