https://docs.opennebula.io/7.2/solutions/ai_factory_blueprints/deployment/Recent content in Deployment onHugoenTue, 21 Oct 2025 00:00:00 +0000https://docs.opennebula.io/7.2/solutions/ai_factory_blueprints/deployment/cd_on-premises/Tue, 21 Oct 2025 00:00:00 +0000https://docs.opennebula.io/7.2/solutions/ai_factory_blueprints/deployment/cd_on-premises/<p><a id="cd_on-premises"></a></p> <p>Machine Learning (ML) training and inference are resource-intensive tasks that often require the full power of a dedicated GPU. PCI passthrough allows a Virtual Machine (VM) to have exclusive access to a physical GPU, delivering bare-metal performance for the most demanding AI workloads.</p> <p>In this guide you will find the details to deploy and configure an AI-ready OpenNebula cloud using the <a href="https://github.com/OpenNebula/one-deploy">OneDeploy</a> tool. It covers the general process for preparing an environment for demanding AI workloads by leveraging PCI passthrough for GPUs like the NVIDIA H100 and L40S.</p>https://docs.opennebula.io/7.2/solutions/ai_factory_blueprints/deployment/cd_cloud/Tue, 21 Oct 2025 00:00:00 +0000https://docs.opennebula.io/7.2/solutions/ai_factory_blueprints/deployment/cd_cloud/<p><a id="cd_cloud"></a> This document describes the procedure to deploy an AI-ready OpenNebula cloud using OneDeploy on a single <a href="https://www.scaleway.com/en/elastic-metal/">Scaleway Elastic Metal</a> bare-metal server equipped with GPUs.</p> <p>The architecture is a converged OpenNebula installation, where the frontend services and KVM hypervisor run on the same physical host. This approach is ideal for demonstrations, proofs-of-concept (PoCs), or for quickly trying out the solution without the need for a complex physical infrastructure.</p> <p>The outlined procedure is based on an instance with NVIDIA L40S GPUs as an example.</p>