Overview
Explore the benefits of GPU utilization in Kubernetes clusters on vSphere infrastructure in this 39-minute conference talk by Steven Wong and Myles Gray from VMware. Learn how to optimize GPU resources across a vSphere cluster, enabling broader access to GPU compute capacity for Kubernetes workloads, even on nodes without physical GPUs. Discover techniques for sharing GPU resources to accelerate AI, deep learning, and inference tasks, potentially reducing costs and improving resource utilization. Follow along with a practical example using TensorFlow workloads on Knative. Gain insights into best practices for hosting Kubernetes on VMware infrastructure and find out how to participate in the Kubernetes VMware User Group for ongoing knowledge sharing and collaboration.
Syllabus
Kubernetes VMware User Group: Using GPUs with K8s on vSphere - Steven Wong & Myles Gray, VMware
Taught by
CNCF [Cloud Native Computing Foundation]