OpenNebula Systems is heading to NVIDIA GTC Berlin 2026, taking place October 20–22, to showcase how organizations can turn NVIDIA-accelerated infrastructure into flexible, multi-tenant AI services. 

As the only European cloud platform fully validated by NVIDIA for AI Cloud, with additional validation for GPU virtualization, OpenNebula provides a production-ready foundation for NVIDIA Cloud Partners, Neoclouds, and AI Factories. At GTC, we’ll show how this foundation supports the complete AI infrastructure lifecycle, from bare-metal provisioning and elastic GPU capacity to Kubernetes, Slurm, and production Token Factory services. 

Find us at Booth 7031 to explore our latest AI Factory demos and solutions. 

NVIDIA AI Factory Demos

A central focus at the OpenNebula booth will be how to turn NVIDIA infrastructure into a complete, multi-tenant AI Factory through a single cloud management platform.

The demo “From Bare Metal to Token Factories: Build Your NVIDIA AI Factory with OpenNebula” follows the full service lifecycle, from automated bare-metal provisioning and infrastructure management to on-demand GPU virtual machines, GPU-enabled Kubernetes clusters, and production Token Factory services.

The demo highlights OpenNebula’s integration with the NVIDIA ecosystem, including GPU passthrough and virtualization, NVIDIA networking and GPU monitoring technologies, and AI software. It shows how Neocloud and AI Factory operators can offer multiple consumption models – from infrastructure capacity to managed inference and token-based AI services – while maintaining centralized control over tenancy, resource allocation, networking, monitoring, and lifecycle management.

The second demo, “Elastic GPU Capacity with NVIDIA Run:ai and OpenNebula,” shows how OpenNebula and NVIDIA Run:ai can transform fixed GPU clusters into elastic capacity services for Neoclouds and AI Factories. GPU-enabled Kubernetes worker nodes and virtual machines can be dynamically provisioned or reclaimed according to workload demand, GPU requirements, quotas, and project-level policies. This enables operators to provide guaranteed GPU capacity while making unused infrastructure available to additional workloads, improving utilization and infrastructure economics.

The third demo, “Elastic Slurm: Scale NVIDIA AI and HPC Clusters on Demand,” shows how OpenNebula can turn NVIDIA GPU infrastructure into elastic Slurm capacity for HPC and AI workloads. Using OneSlurm, operators can provision isolated Slurm environments and dynamically add or release GPU worker nodes as demand changes, allowing projects to burst beyond fixed cluster capacity while returning idle resources to the shared infrastructure pool. The demo illustrates how the same accelerated infrastructure can support traditional HPC, distributed AI training, and other batch workloads with centralized tenancy, quota, and lifecycle control.

Learn more about OpenNebula’s NVIDIA AI Cloud Ready validation and the validated architecture behind production AI Factories here.

More AI Factory Demos with Our Technology Ecosystem

In addition to the core OpenNebula demos, we plan to showcase additional AI Factory use cases in collaboration with our technology ecosystem. These demonstrations will highlight how OpenNebula integrates with complementary technologies across storage, networking, accelerated computing, data platforms, AI software, and infrastructure services.

One example is “From AI Data to GPU Inference in Minutes,” which shows how OpenNebula, NVIDIA, and NetApp can combine persistent AI data with on-demand GPU infrastructure. NetApp ONTAP provides centralized, persistent model and dataset repositories, while OpenNebula delivers the cloud management and orchestration layer for tenant isolation, networking, GPU allocation, and Kubernetes lifecycle management. Through OneKS and NetApp Trident, Kubernetes clusters can access existing AI data and launch NVIDIA NIM-powered inference services in minutes, without duplicating or moving the underlying datasets.

The broader objective is to demonstrate how Neocloud and AI Factory operators can combine OpenNebula with best-of-breed ecosystem technologies while maintaining a consistent cloud management and orchestration layer across bare metal, GPU virtual machines, Kubernetes, Slurm, and Token Factory services.

Building an AI Factory? Explore how OpenNebula can help you turn NVIDIA infrastructure into production-ready AI services here.

Meet OpenNebula at GTC Berlin

Find the OpenNebula Systems team at Booth 7031 to explore these demos and discuss how NVIDIA infrastructure, open cloud management, Kubernetes, data platforms, and GPU orchestration can come together to build production-ready AI environments.

Whether you are building an AI Factory, operating a Neocloud, or looking to bring enterprise AI workloads into production, the team will be available to discuss your infrastructure and requirements.

Visit Booth 7031 to see how OpenNebula and its technology ecosystem can turn accelerated infrastructure into consumable AI services.