Mirantis IaaS Technology Stack for GPU Infrastructure Management

Explore the infrastructure-as-a-service technology stack that powers Mirantis k0rdent AI in this 23-minute conference presentation. Learn how Mirantis designed their IaaS layer to manage bare metal servers, networking, and storage resources specifically optimized for GPU operations, removing traditional friction points that slow down AI infrastructure deployment. Discover the automated provisioning capabilities that handle bare metal setup, integrate with heterogeneous server and network vendors, and implement governance for tenancy and workload isolation. Understand how validated GPU drivers reduce compatibility risks and accelerate time-to-production for AI workloads. Watch a live demonstration showing the creation of a new GPU cluster through the k0rdent AI interface, including automatic hardware discovery, network overlay configuration, and compute resource assignment. See how administrators can monitor GPU usage at the device level with real-time allocation and health data tracking. Gain insights into how this focused approach to infrastructure automation enables smaller teams to manage large-scale GPU environments that previously required extensive staffing, while maintaining vendor neutrality and reducing integration complexity for organizations scaling their AI operations.