How to Deploy a High-performance Distributed AI Training Cluster with NVIDIA GPUs and KVM (Presented by IBM)
, Research Scientist, IBM Research
, Principal Technologist/Solutions Architect, NVIDIA
Are your bare-metal AI systems too rigid for your use cases? Would you like to securely share them among multiple users using virtual machines like a cloud provider? This technical talk describes how to enable GPUs in virtual machines and achieve optimal performance.