Multi-Cloud Machine Learning Data and Workflow with Kubernetes
summary
Describes using Kubernetes to build a multi-cloud machine learning platform for autonomous-vehicle workloads, covering management of training data across on-prem and public clouds with different GPUs and network interfaces (Infiniband, RoCE). Explains how Kubernetes is used to handle multi-user and gang scheduling and to support heterogeneous hardware in those environments.
AI-written summary · about
related talks
36mMulti-Cloud Machine Learning Data and Workflow with Kubernetes
06mKubernetes Love Machine Learning, Even on Private Cloud
46mMachine Learning with Kubernetes
31mBuilding and Managing a Centralized ML Platform with Kubeflow at...
53mUsing Kubernetes for Machine Learning Frameworks
33mKubernetes-Driven Multi-Cloud Deployments: Maximizing Scalability and Flexibility
31mKubernetes For GPU Powered Machine Learning Workloads In Academia
26mMachine Learning Meets K8s