NVIDIA/cuml

Health v2 · maintenance only

Adoption not part of the score

5292 stars · 683 forks observed · 2026-09-30

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

NVIDIA cuML is a GPU-accelerated machine learning library offering scikit-learn-style estimators that run on NVIDIA GPUs via CUDA. It also provides cuml.accel, a drop-in accelerator for existing scikit-learn, UMAP, and HDBSCAN code, plus multi-GPU support through Dask.

Use cases

• speed up scikit-learn workflows on gpu
• run clustering and regression on nvidia gpus
• accelerate umap and hdbscan without changing code
• scale machine learning across multiple gpus
• gpu-accelerated nearest neighbor search
• drop-in replacement for sklearn estimators

When to choose

• you have nvidia gpus and want large speedups on sklearn-style workloads
• you want to accelerate existing scikit-learn code with minimal changes
• you need multi-gpu or multi-node distributed machine learning with dask

When to avoid

• you have no nvidia gpu hardware
• you rely on estimators not yet supported by cuml.accel and need guaranteed cpu fallback behavior
• you need a pure-cpu portable solution

Facets

library · maturity active

machine-learning data-science gpu-computing machine-learning data-science gpu-computing python cuda scikit-learn rapids dask acceleration gpu linux docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="NVIDIA/cuml")