Event Details

December 4, 2026

Join Zoom Webinar

Host: Steve Crago
POC: Amy Kasmir

This seminar reviews recent advances in analog, digital, and probabilistic compute-in-memory (CIM) using silicon CMOS technologies and emerging memory devices. We begin with multiple generations of silicon-proven prototype chips, featuring current-domain ACIM based on resistive random-access memory (RRAM) fabricated in TSMC’s 40 nm process and ferroelectric field-effect transistors (FeFET) implemented on GlobalFoundries’ 28 nm FeFET platform. We also introduce a more energy-efficient charge-domain ACIM approach enabled by ferroelectric nonvolatile capacitors (nvCaps) on GlobalFoundries 28nm FeFET platform. We examine the challenges of ACIM, particularly inference accuracy degradation arising from process, voltage, and temperature (PVT) variations, and discuss corresponding mitigation techniques. These challenges motivate more robust digital CIM (DCIM) architectures, exemplified by FeFET-based divider bit cells. Finally, we explore how stochasticity in memory bit cells can enable probabilistic CIM (pCIM) for Ising machines that solve combinatorial optimization problems, highlighting designs incorporating noisy static random-access memory (SRAM) and magnetic random-access memory (MRAM) taped out on TSMC’s 28 nm and 16 nm platforms.

Speaker Bio

Shimeng Yu is the endowed Dean’s Professor of Electrical and Computer Engineering at the Georgia Institute of Technology. He received a PhD degree from Stanford University in 2013. He is an elevated IEEE Fellow for contributions to non-volatile memories and in-memory computing. Prof. Yu’s general research interests are emerging devices and 3D integrated circuits for AI hardware. Among Prof. Yu’s recent honors, he was a recipient of Intel Outstanding Researcher Award in 2023, and Semiconductor Research Corporation (SRC) Rama Divakaruni Technical Excellence Award in 2025. Prof. Yu’s 500+ journal/conference publications received more than 45,000 citations (Google Scholar) with H-index 93. For the past decade, Prof. Yu has served as technical program committee and frequent short course/tutorial speakers for the leading conferences in the field, including IEEE International Electron Devices Meeting (IEDM), IEEE Symposium on VLSI Technology and Circuits, IEEE International Memory Workshop (IMW), and IEEE International Reliability Physics Symposium (IRPS), etc. He also has served as an editor for IEEE Electron Device Letters (EDL).