Boston
May 8-11, 2017
Canonical - Using Kubernetes and OpenStack for CPU and GPU Intensive Workloads
It is commonly thought that OpenStack and/or Kubernetes are not appropriate enviropnments for CPU intensive workloads of the types of application that make use of GPUs. In this session Marco will show how applications such as tensorflow or video rendering, managed by Kubernetes running in OpenStack not only delivers performance but can actually improve the overall compute throughput.
Monday, May 8, 12:05pm-12:45pm (4:05pm - 4:45pm UTC)
Difficulty Level: Intermediate
Principal Architect
Marco Ceppi is a DevOps Engineer helping build the future of clouds. He's been managing deployments at scale for 10 years on both physical and virtual machines. In his spare time he takes delight in pushing the limits of technology and finding ways to improve on performance, reliability, and awesomeness. FULL PROFILE
Canonical, Principal Engineer
James Page is an Ubuntu and Debian Developer and previous PTL of the OpenStack Charms project.
James is currently focussed on distribution and deployment of OpenStack on Ubuntu and currently works for Canonical as Technical Architect of the OpenStack Engineering team. FULL PROFILE
Download App
