IBM makes the case for doing deep learning on Kubernetes



Jason McGee. Source: IBM

In this episode of the ARCHITECHT Show, IBM’s Jason McGee (VP and CTO, IBM Cloud) and Ruchir Puri (chief architect, IBM Watson) explain why Kubernetes can be an ideal platform for running deep learning and AI workloads. Among other topics, the two go into detail on an open source project called FfDL (Fabric for Deep Learning), how businesses are taking advantage of advances in AI and application architecture, and ever-expanding set of hardware platforms targeting AI.

Scroll to the bottom (or click here) for links to listen to the podcast pretty much everywhere else you might want to.

How to listen to the ARCHITECHT Show everywhere else

Source: Deep Learning on Medium