GPU Sharing Done Right - Secrets, Security, and Scaling with Vault and vCluster

Learn how to build a secure, scalable multi-tenant Kubernetes infrastructure for GPU sharing in AI workloads through this 28-minute conference talk. Discover the challenges of balancing performance and security when multiple teams need access to shared GPU resources, and explore practical solutions for creating strong isolation and access control. Watch a comprehensive demonstration of integrating HashiCorp Vault for secrets management and dynamic policy enforcement, combined with vCluster technology to provide teams with the experience of having their own dedicated cluster while efficiently pooling resources behind the scenes. Gain insights into best practices for implementing this powerful AI infrastructure model that enables organizations to maximize GPU utilization while maintaining enterprise-grade security and tenant isolation. The presentation covers real-world implementation strategies, security considerations, and scaling approaches that allow teams to run GPU workloads with confidence in a shared environment.