Twenty years ago, the end of Dennard scaling triggered a paradigm shift in computer architecture design: today, system designers must consider a broad array of architectures and components to meet escalating computational demands. Co-design methodologies have emerged to address this challenge, enabling designers to interact across traditional design hierarchies and explore novel trade-offs between previously isolated domains.

In this seminar, we present SystemFlow, a framework that systematically describes diverse systems and enables quantitative comparison of key performance metrics—such as the power savings achieved in data center processing systems by introducing on-sensor data reduction. Furthermore, these estimates showing the importance of early-stage data reduction inspired further conversations between devices and architectures, leading us to develop an analog neuromorphic system which provides in-sensor data filtering without the use of digital quantization, leading to improvements in filtering performance. Currently, we are extending these and other techniques to applications in the X-ray sciences and beyond to enable cutting-edge computational systems for scientific disciplines.

Dr. Wilkie Olin-Ammentorp is a post-doctoral researcher in the Mathematics & Computer Science Division of the Argonne National Laboratory. He completed his PhD in Nanoengineering at the State University of New York Polytechnic Institute in 2019 and went on to a post-doctoral appointment at the University of California, San Diego in the Department of Medicine before joining Argonne in 2022. His research focuses on designing, simulating, and benchmarking novel computing architectures which combine the strengths of modern computing systems as well as principles from biological computation.