Skip to navigation

Source Library

View as Markdown

This index favors current product documentation, Enterprise Reference Architectures, security guidance, published reference implementations, and publisher-maintained ecosystem links. Last systematic review: November 12, 2026.

Diagram Assets

Download the static architecture PNG

Download the scalable stack SVG

Architecture and Infrastructure

Use these sources for infrastructure architecture, compatibility, and observability guidance:

SubjectPrimary Source
Enterprise infrastructure design pointsNVIDIA Enterprise Reference Architectures
Release-specific infrastructure compatibilityNVIDIA AI Enterprise Infrastructure Support Matrix
NVIDIA AI EnterpriseNVIDIA AI Enterprise documentation
Full-stack observabilityAI Factory Observability Guide
NVIDIA DSX navigation referenceNVIDIA DSX documentation

AI Platform

Use these sources for current NVIDIA AI Platform product guidance:

Cloud-Native Platform

Use these sources for current NVIDIA Cloud-Native Platform guidance:

SubjectPrimary Source
GPU OperatorNVIDIA GPU Operator documentation
Network OperatorNVIDIA Network Operator documentation
NIM OperatorNVIDIA NIM Operator documentation
GPU telemetry and diagnosticsNVIDIA DCGM

Security and Capability Designs

Use these sources for security guidance and published Capability Reference Designs:

SubjectPrimary Source
NVIDIA Trusted Computing SolutionsNVIDIA Trusted Computing Solutions library
NVIDIA AttestationNVIDIA Attestation documentation
Secure AI operationsNVIDIA Secure AI Operations Guide
NVIDIA Open Agent Safety PlatformNVIDIA Open Agent Safety Platform
Open Agent Safety Platform architectureContinuous in-silicon agent monitoring
NVIDIA OpenShellOpenShell Developer Guide
Secure Agent Workspace with OpenShellBuilding secure agent workspaces with NVIDIA OpenShell Reference Architecture
Confidential inference on virtual machinesDeploying Proprietary Models Securely with NVIDIA Confidential Computing on Self-Hosted Virtual Machines
Confidential inference on KubernetesDeploying Proprietary Models Securely with NVIDIA Confidential Computing on Self-Hosted Kubernetes
Inference Hub architecture and adoptionAI Factory Inference Hub Reference Design
Inference Hub component examplesPublished component mapping
Inference Hub on Red Hat AI Factory with NVIDIAPublished Red Hat example
SAW on Red Hat OpenShift VirtualizationSAW on Red Hat OpenShift Virtualization
SAW on Microsoft AzureSAW on Microsoft Azure

Inference Hub source review: November 18, 2026. The integration guide explains its relationship to the core, SAW, and Industry Profiles. Detailed architecture, request flow, security, operations, adoption, and implementation guidance remain in the published Inference Hub design.

Agent tooling, workspace, and confidential-inference source review: November 13, 2026.

Government and Regulated-Environment Sources

Use these sources for Government and regulated-environment requirements:

SubjectPrimary Source
Cryptographic module requirementsFIPS 140-3
Cryptographic module validation statusCMVP
Security and privacy control catalogNIST SP 800-53 Rev. 5
Control baselinesNIST SP 800-53B
Control assessment methodologyNIST SP 800-53A Rev. 5
Risk management and authorizationNIST SP 800-37 Rev. 2
Product hardening assetsDISA STIG Downloads
Government profile decisionsGovernment Industry Profile

Ecosystem Source Lineage

The Ecosystem Solution Catalog consolidates the software-partner sections of two source publications reviewed on November 12, 2026. The sources are the Enterprise AI Factory Design Guide and the Government reference-design publication. The catalog preserves publisher URLs and source-publication labels while deduplicating organizations across categories and profiles. Use it for solution discovery and architecture planning, then verify current technical and support requirements with each publisher.

  • Core source inventory: 58 named offerings across AI and AgentOps, observability, security, integrated software platforms, cloud-native platforms, and storage.
  • Government source inventory: 22 entries mapped onto the same stable facets and merged into existing partner records where applicable.
  • Open an offering to read its original published description. Where both publications provide descriptions, the Core and Government text appears separately.

OEM AI Factory Offerings

The OEM AI Factory Offerings page links directly to publisher-owned full-stack offerings from Cisco, Dell Technologies, HPE, Lenovo, and Supermicro. It supports discovery without reproducing orderable configurations, bills of materials, compatibility matrices, commercial terms, or support claims.

  • Prefer the most specific authoritative page that supports the nearby statement.
  • Prefer stable product roots. Use release-specific links only when no stable root exists.
  • Keep time-sensitive support and compatibility matrices in their owning product documentation.
  • Do not link the replacement site back to the two source publications it consolidates.
  • Record the review date when a systematic source review changes this index.