Skip to main content
Ctrl+K
RAPIDS Deployment Documentation - Home RAPIDS Deployment Documentation - Home

RAPIDS Deployment Documentation

nightly stable
  • Docs Home
  • GitHub
RAPIDS Deployment Documentation - Home RAPIDS Deployment Documentation - Home

RAPIDS Deployment Documentation

nightly stable
  • Docs Home
  • GitHub

Table of Contents

  • Local
    • Custom RAPIDS Docker Guide
  • Cloud
    • NVIDIA Cloud Platforms
      • NVIDIA Brev
    • Amazon Web Services
      • Elastic Compute Cloud (EC2)
      • EC2 Cluster (via Dask)
      • AWS Elastic Kubernetes Service (EKS)
      • Elastic Container Service (ECS)
      • SageMaker
    • Microsoft Azure
      • Azure Virtual Machine
      • Azure Kubernetes Service
      • Azure VM Cluster (via Dask)
      • Azure Machine Learning
    • Google Cloud Platform
      • Compute Engine Instance
      • Vertex AI
      • Google Kubernetes Engine
      • Dataproc
    • IBM Cloud
      • Virtual Server for VPC
  • HPC
  • Platforms
    • NVIDIA AI Workbench
    • Kubernetes
    • Kubeflow
    • KServe
    • Coiled
    • Databricks
    • RAPIDS on Google Colab
    • Snowflake
    • Modal
  • Tools
    • dask-cuda
    • Dask Operator
    • Dask Helm Chart
  • Workflow Examples
    • Scaling up Hyperparameter Optimization with Kubernetes and XGBoost GPU Algorithm
    • Scaling up Hyperparameter Optimization with Multi-GPU Workload on Kubernetes
    • Getting Started with Optuna and RAPIDS for HPO
    • Running RAPIDS Hyperparameter Experiments at Scale on Amazon SageMaker
    • Deep Dive into Running Hyper Parameter Optimization on AWS SageMaker
    • Multi-node Multi-GPU Example on AWS using dask-cloudprovider
    • Autoscaling Multi-Tenant Kubernetes Deep-Dive
    • HPO with dask-ml and cuml
    • Train and Hyperparameter-Tune with RAPIDS on AzureML
    • Perform Time Series Forecasting on Google Kubernetes Engine with NVIDIA GPUs
    • HPO Benchmarking with RAPIDS and Dask
    • Training XGBoost with Dask RAPIDS in Databricks
    • Multi-Node Multi-GPU XGBoost Example on Azure using dask-cloudprovider
    • Measuring Performance with the One Billion Row Challenge
    • Getting Started with cudf.pandas and Snowflake
    • Getting Started with cuML’s accelerator mode (cuml.accel) in Snowflake Notebooks
    • Accelerating data analysis using cudf.pandas
    • Deploying End-to-End Kafka Streaming SI Detection Pipeline with cuDF, Morpheus, and Triton on EKS
    • Orchestrating a Fraud Detection Model Lifecycle with cuDF, Prefect, MLflow, and Triton
    • GPU-Accelerated Land Use Land Cover Classification
    • HPO for Random Forest with Ray Tune and cuML
  • Guides
    • Multi-Instance GPU (MIG)
    • Building RAPIDS containers from a custom base image
    • How to Setup InfiniBand on Azure
    • Does the Dask scheduler need a GPU?
    • GPU optimization for the Dask scheduler on Kubernetes
    • Colocate Dask workers on Kubernetes while using nodes with multiple GPUs
    • Caching Docker Images For Autoscaling Workloads
  • NVIDIA NIM Microservices
  • Developer
    • Continuous Integration
      • GitHub Actions
  • Local

Local#

Choose your preferred installation method for running RAPIDS

conda

Install RAPIDS using conda

https://docs.rapids.ai/install#conda

Docker

Install RAPIDS using Docker

Custom RAPIDS Docker Guide

pip

Install RAPIDS using pip

https://docs.rapids.ai/install#pip

WSL2

Install RAPIDS on Windows using Windows Subsystem for Linux version 2 (WSL2)

https://docs.rapids.ai/install#wsl2

previous

NVIDIA RAPIDS Deployment Documentation

next

Custom RAPIDS Docker Guide

NVIDIA NVIDIA
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2026, NVIDIA.