Coordinated Design of Workloads and Systems via Machine Learning

Abstract

Rapidly advancing AI and ML have the potential to revolutionize the design of increasingly complex computer systems. With this vision, we are building an open source end-to-end automated framework for the coordinated design of workloads and systems, intelligently managed through ML techniques. The aim of the framework is two-fold: optimize the design of workloads to efficiently evaluate the computer system and optimize the design of the computer system to match the needs of the domain / workloads.