NVIDIA Confidential Containers Architecture#

This is documentation for NVIDIA’s implementation of Confidential Containers including reference architecture information and supported platforms.

Overview

Start here to review the reference architecture, use cases, and software components.

NVIDIA Confidential Containers Reference Architecture
Supported Platforms

Learn about the validated hardware, OS, and component versions.

Supported Platforms
Prerequisites

Hardware, BIOS, and Kubernetes cluster requirements.

Prerequisites
Deploy Confidential Containers

Install Kata Containers and the NVIDIA GPU Operator on Kubernetes.

Deploy Confidential Containers
Run a Sample Workload

Verify your deployment by running a GPU workload in a confidential container.

Run a Sample Workload
Managing the CC Mode

Set the confidential computing mode on NVIDIA GPUs at cluster or node level.

Managing the Confidential Computing Mode
Multi-GPU Passthrough

Configure multi-GPU passthrough for NVSwitch-based HGX systems.

Configuring Multi-GPU Passthrough Support
Image Pull Timeouts

Tune image pull timeouts for large container images in confidential VMs.

Configure Image Pull Timeouts
Attestation

Remote attestation, Trustee, and the NVIDIA verifier for GPU workloads.

Attestation
Release Notes

New features and known issues for each release.

Release Notes
Licensing

Licensing information for Confidential Containers documentation.

Licensing