Accelerating the AI Journey: SUSE AI Factory with NVIDIA

The journey from a successful AI prototype to a stable, scalable production environment is often where innovation hits a wall. For many organizations, the “hidden technical debt” of AI—managing GPU drivers, securing model supply chains, and bridging the gap between a data scientist’s laptop and a hardened data center—remains a significant barrier.

To solve this, SUSE  is bringing to market SUSE AI Factory with NVIDIA. This integrated solution will provide an end-to-end “assembly line” for the modern enterprise, combining the mission-critical reliability of the SUSE AI stack with the high-performance capabilities of NVIDIA AI Enterprise.

The Architecture: A Unified Stack for Tomorrow’s Innovation

The SUSE AI Factory with NVIDIA is being co-designed as an easy to use but meticulously integrated platform providing a jointly engineered stack that spans from the underlying hardware to the AI apps and pre-validated blueprints.  This full stack approach ensures that every layer is optimized for performance and reliability.  The core software components of this platform are:

• Foundation & Orchestration: Built on SUSE Linux Enterprise Server (SLES) and orchestrated by SUSE Rancher Prime (utilizing RKE2 or K3s) with NVIDIA Run:ai, providing an industry leading environment for containerized AI workloads that come with the validated GPU drivers and operators.
• NVIDIA AI Enterprise Integration: Deep integration with NVIDIA AI Enterprise  gives teams access to a production-grade suite, including NVIDIA NIM™ (microservices for optimized inference), open Nemotron models, NVIDIA NeMo™ for building and customizing generative AI models, and the NVIDIA OpenShell runtime for deploying more secure, policy-governed autonomous AI agents.
• Automated Infrastructure: The solution includes all necessary AI infrastructure operators, such as the NVIDIA GPU Operator, NIM Operator, and Network Operator, to automate the lifecycle of specialized AI hardware.
• Enterprise-Grade Integrity: The platform features integrated governance, including security and observability to deliver real-time insights into GPU utilization and ensure production models remain performant and safe.
• Curated Software Supply Chain: By including the SUSE Application Collection, the solution provides users with a curated set of hardened, pre-signed container images, ensuring that open-source tools meet strict enterprise security standards.

Redefining Enterprise Simplicity

One of the core pillars of the SUSE AI Factory is Enterprise Simplicity. We are moving away from fragmented, complex tools toward a seamless, UI-driven deployment model that lowers the barrier to entry for AI innovation.

• Flexible Management Styles: We offer ultimate flexibility by supporting both ClickOps and GitOps workflows. Administrators can use the intuitive user interface for ClickOps simplicity, making the deployment of complex AI applications extremely easy. For scale and automation, platform engineers can leverage Fleet for GitOps, managing deployments via code to ensure AI workloads are reproducible, auditable, and easily scalable across massive environments.
• Blueprint-Driven Deployment: Instead of starting from scratch, users can leverage and build upon or customize pre-validated, opinionated turnkey blueprints—such as the NVIDIA Retrieval-Augmented Generation (RAG) and the NVIDIA AI-Q Research Assistant—to move from concept to production in days rather than months.
• Unified Management Plane: Whether you are deploying in a central data center, within a virtual private cloud,or at the far edge to power physical AI in a smart factory or a robotic warehouse, the solution provides a consistent management experience across all environments.

From Developers to Production Line

The SUSE AI Factory aims to bridge the gap between development and operations through two interconnected environments:

1. Rapid iteration for developers

We enable developers to mirror the production stack locally and then help them to promote their applications and blueprints

• Frictionless Setup: Developers can start coding immediately with the full NVIDIA AI Enterprise stack in a local sandbox.
• Agentic Innovation: The integration of NVIDIA OpenShell will allow developers to build and deploy secure, on-premises autonomous AI agents, keeping proprietary logic within private infrastructure.

2. The Production Line: Scalable Deployment

Once an application is refined, the platform will enable a seamless promotion to production clusters.

• Industrial-Scale GitOps: For DevOps-oriented teams, the solution leverages Rancher Continuous Delivery (Fleet) to manage Kubernetes-based deployments at scale, ensuring consistent, version-controlled environments via declarative configurations.

The Path Forward

The SUSE AI Factory with NVIDIA is designed to meet the rigorous demands of the modern enterprise:

1. Sovereign Design: Purpose-built for organizations with strict data sovereignty requirements, ensuring total control over data, models, and intellectual property.
2. Unified Enterprise Support: SUSE will provide a single point of accountability, offering comprehensive support across the entire stack, including L1 and L2 support for NVIDIA components.

As we look toward the launch, the SUSE AI Factory with NVIDIA represents our commitment to turning complex AI experimentation into a secure, sovereign, and scalable reality.  Learn more at suse.com/products/ai/factory-with-nvidia.