NVIDIA/cudf-spark

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

1012 stars · 311 forks observed · 2026-10-20

What it is AI-extracted, prompt v1, taxonomy v1, 2026-09-07, confidence 0.95

A plugin library for Apache Spark from NVIDIA that leverages GPUs via the cuDF (CUDA DataFrame) library to accelerate Spark SQL and DataFrame processing. It was formerly known as the RAPIDS Accelerator for Apache Spark.

Use cases

• accelerate Apache Spark SQL and DataFrame jobs with GPUs
• speed up big data ETL pipelines without rewriting Spark code
• run cost-performance tuning for GPU-accelerated Spark clusters
• zero-copy data transfer from Spark into GPU machine learning libraries like XGBoost
• check operator compatibility to keep results bit-for-bit identical with CPU Spark

When to choose

• you run large Spark SQL/DataFrame workloads and have NVIDIA GPUs available
• you want drop-in acceleration for existing Spark jobs without code changes
• you need identical results to vanilla Spark while gaining GPU performance

When to avoid

• you have no NVIDIA GPU hardware or cloud GPU instances
• your workloads use Spark operators not covered by the plugin's compatibility matrix
• you only run small jobs where GPU acceleration overhead outweighs gains

Facets

plugin · maturity active

gpu-computing data-science etl database big-data data-science gpu-computing analytics jvm apache-spark rapids cudf cuda dataframe sql-acceleration gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="NVIDIA/cudf-spark")