Declarative Device Virtualization - Orchestrating GPUs and Hardware in Cloud Native Environments

Explore how declarative pipelines revolutionize GPU and hardware virtualization within Kubernetes environments in this 34-minute conference talk from the Linux Foundation. Address the challenges of managing specialized hardware resources in cloud-native applications and learn to orchestrate virtualized devices with ease. Discover how to define device configurations using YAML, deploy virtualized GPUs, FPGAs, and other accelerators into Kubernetes clusters, and automate complex hardware interactions. Examine practical examples of using declarative pipelines for AI/ML workloads, edge computing, and high-performance computing (HPC) scenarios. Gain insights into building scalable, resilient, and adaptable device virtualization solutions with specific focus on GPU management and optimization techniques. Learn from Google experts Samrat Priyadarshi and Anmol Krishan Sachdeva as they demonstrate how to unlock the full potential of hardware resources in cloud-native environments through declarative approaches to device virtualization.