About

Andreas Gerstlauer is a Cullen Trust for Higher Education Endowed Professor and Associate Chair for Academic Affairs in the Electrical and Computer Engineering Department at The University of Texas at Austin. He received his PhD in Information and Computer Science from the University of California in 2004. Before joining UT Austin in 2008, he was an Assistant Researcher at the Center for Embedded Computer Systems at UC Irvine, where he led a research group that developed electronic system-level design tools. Dr. Gerstlauer is a co-author of 3 books and more than 150 conference and journal publications. His work was recognized with the 2024 HASP, 2021 MLCAD, 2016 DAC and 2015 SAMOS best paper awards, several best paper nominations from DAC, DATE, FCCM and HOST conferences, as a 2021 IEEE HSTTC Top Pick in Hardware and Embedded Security, as one of the most influential contributions in 100 years at DATE in 2008, and as recipient of a 2016-2017 Humboldt Research Fellowship. He has served as a Senior, Associate, and Special Issue Editor for the ACM TECS and TODAES journals, as well as General or Program Chair for major international conferences such as ESWEEK, MEMOCODE, CODES+ISSS, and SAMOS. His research interests include system-level design automation, system modeling, design languages and methodologies, and embedded hardware and software synthesis.

Abstract

System-Level Design of Heterogeneous, Accelerator-Rich System Architectures

Traditional semiconductor scaling reaching physical and technological limits has driven the need for increasing specialization and heterogeneity in computer system design. The use of custom compute units, such as GPUs or dedicated hardware accelerators, is commonplace across many application domains, including machine learning (ML). At the same time, increasing heterogeneity creates challenges for designing and managing such complex systems and chips, where the use of ML techniques has seen a lot of excitement and promise. In this talk, Andreas Gerstlauer will discuss recent work on both system design for ML and ML for system design.

In heterogeneous, accelerator-rich system architectures, data movement is rapidly becoming the bottleneck, and even on-chip architectures must be treated as distributed systems. The lecture will cover architectural support for data movement orchestration, including near-memory computing, as well as emerging superconducting and neuromorphic computing paradigms to break through physical power and memory walls. On the design side, applications of ML to system-level modeling, design space exploration, and runtime management will be presented. This includes ML-based approaches for cross-layer, cross-platform, and cross-temporal performance/energy prediction, as well as applications of predictive models to application/architecture co-optimization and proactive runtime resource management of heterogeneous systems.

About Current Trends in Computer Science

This lecture is part of the Current Trends in Computer Science Lecture Series by the TU Wien Informatics Doctoral School, where renowned Guest Professors hold public lectures every semester. If you are studying with us, the lecture series can be credited as an elective course for students of master’s programs of computer science: 195.072 Current Trends in Computer Science. Additionally, you can join courses held by this year’s Guest Professors of our doctoral colleges and the TU Wien Informatics Doctoral School.

Curious about our other news? Subscribe to our news feed, calendar, or newsletter, or follow us on social media.