Sustainable Compute for the Broad AI Era

Wednesday, November 30, 2022 | 12:00 – 1:00pm ET,
MIT Building 34-401A, Grier Room A
Speaker: John Rozen, IBM Research

View the recording of this Joint MTL/AI Hardware Program Seminar here.

Abstract: We are witnessing a transition in AI. Narrow systems able to execute specific tasks in a single domain are giving way to broad AI that learns more generally and can work across domains and problems. We need more innovation in hardware and infrastructure to power and deploy the next generation of AI models. Continuing to rely on conventional infrastructure and processors is simply not scalable — or sustainable.

In this talk, we will review the type of AI models driving this trend and the emerging approaches to tackle the sustainability issue at different levels of the computing stack. The hardware and software opportunities indeed are everywhere; from redefining the fundamental compute unit, to shrinking technology nodes, optimizing design of cores & architecture, building large systems, and integrating into the cloud fabric.

Speaker Bio: John Rozen, PhD, a Senior Research Scientist serving as Program Director of the IBM Research AI Hardware Center located in NY. Rozen was responsible for establishing a SiC MOS Research Program at the Central Research Institute of the Electric Power Industry in Japan. At IBM, he is the technical lead for a high-k/metal-gate joint development program dedicated to CMOS technology development for the 7nm node and beyond. Rozen also manages the materials team developing novel synaptic elements towards building accelerators for machine learning workloads as part of the IBM AI Hardware Center. Rozen serves on the Scientific Advisory Boards of academia-industry consortia, including the DARPA- & NSF-funded SRC nCORE program. Rozen received a BS and an MS in Physics from the Université Libre de Bruxelles, Brussels, Belgium, and an MS and PhD in Materials Science from Vanderbilt University.