Beyond von Neumann Architectures for Next-Generation Sensing and Computing

Abstract: Conventional digital architectures increasingly waste energy and bandwidth moving data among sensors, memory and processors.

This talk examines mixed-signal and non-von Neumann approaches that attack this bottleneck by exploiting memory locality, circuit dynamics and device physics more directly. I will discuss in- and near-memory computing, memory-centric architectures in Complementary Metal-Oxide-Semiconductor (Static Random Access Memory), Ferroelectric Field Effect Transistors and Resistive Random-Access Memory, and selected dynamical computing approaches, with examples spanning sensor interfaces, adaptive signal processing and neuromorphic and intelligent edge systems.

The focus is not only on potential gains in energy efficiency and compute density, but also on the constraints that determine whether those gains survive at the system level: variation, analog non-idealities, limited precision, programmability and peripheral overheads. A central theme is that useful systems require principled algorithm–architecture–circuit co-design rather than isolated device- or circuit-level innovation.

Series: The Argonne Microelectronics Institute colloquium series invites leaders in the field from academia, national labs and industry to present their forefront research and perspectives. Join us in-person or online to learn and connect.