Embedded Files
Date:
Sunday, March 22, 2026
Time:
13:30 - 18:00 Afternoon Session (15:30-16:00 Coffee Break)
Location:
Room Ft. Duquesne
The Landing Hotel, 757 Casino Dr. Pittsburgh, PA 15212, Pittsburgh
Schedule:
13:30 - 13:45 Opening Remarks
13:45 - 14:45 Keynote: Bit-Stream Computing Without Compromise: Accuracy, Robustness, and Energy Efficiency in the
Post-Random Era
Dr. M. Hassan Najafi
Case Western Reserve University
14:45 - 15:00 From Retina to Silicon: The IRIS Framework for Bio-inspired Visual Intelligence
Subhradip Chakraborty, Shay Snyder*, Maryam Parsa*, Gregory Schwartz^, Akhilesh Jaiswal
University of Wisconsin Madison, George Mason University*, Northwestern University^
15:00 - 15:15 Agraph: A unified Graph Representation for At-Will Simulation of Emerging Stacks
Daniel Price, Prabhu Vellaisamy*, Patricia Gonzalez-Guerrero^, George Michelogiannakis^, John Paul Shen*, Di Wu
University of Central Florida, Carnegie Mellon University*, Lawrence Berkeley National Labs^
15:15 - 15:30 Unary is All You Need: Exploiting Unary Computing in Efficient Hyperdimensional Intelligence
Mehran Moghadam, Sercan Aygun*, Faeze Sadat Banitaba*, M. Hassan Najafi
Case Western Reserve University, University of Louisiana at Lafayette*
15:30 - 16:00 Coffee Break
16:00 - 16:15 PRISM: Posit Representation In Stochastic Multiplication
Zhaojun Ni, Ruyue Yan, Honglan Jiang*, Siting Liu
ShanghaiTech University, Shanghai Jiao Tong University*
16:15 - 16:30 SCoRe: Stochastic Compressed Representations
Paridhi Gupta, Zhewen Pan, Hsuan Hsiao, Joshua San Miguel
University of Wisconsin - Madison
16:30 - 16:45 Beyond Synthetic Randomness: Stochastic Computing via True Randomness
Mehran Moghadam, Shelby Williams, Abu Kaisar Mohammad Masum*, Sercan Aygun*, M. Hassan Najafi
Case Western Reserve University, University of Louisiana at Lafayette*
16:45 - 17:00 In-Memory Stochastic Division for Efficient Retinex Image Enhancement
Farzad Razi, Mehran Moghadam*, M. Hassan Najafi*, Sercan Aygun^, Marc Riedel
University of Minnesota, Case Western Reserve University*, University of Louisiana at Lafayette^
17:00 - 17:15 Mugi: Value Level Parallelism For Nonlinear Operations in LLMs
Daniel Price, Prabhu Vellaisamy*, John Paul Shen*, Di Wu
University of Central Florida, Carnegie Mellon University*
17:15 - 18:00 Miscellaneous (Brainstorming, Roadmap Discussion, etc.)
Keynote
Title: Bit-Stream Computing Without Compromise: Accuracy, Robustness, and Energy Efficiency in the Post-Random Era
Abstract:
Stochastic computing (SC) has long promised cost-efficient and noise-tolerant hardware by representing data as probability-encoded bit-streams. However, its practical adoption has been constrained by the high cost of bit-stream generation, uncontrolled randomness, correlation sensitivity, and excessive data movement. This keynote presents a unified perspective on how recent advances overcome these limitations and enable bit-stream computing without compromise. We begin by rethinking entropy itself. Instead of relying solely on pseudo-random sources, low-discrepancy sequence generation—such as lightweight powers-of-two LD generators—enables faster convergence, higher accuracy, and significantly reduced bit-stream generation cost. We introduce Stochastic Intermittent Computing, where progressive precision and noise tolerance eliminate checkpointing overhead and enable always-on sensing under unstable energy conditions. Finally, we discuss fully in-memory stochastic architectures using emerging devices such as ReRAM, where random generation, computation, and conversion are performed in place, drastically reducing data movement while exploiting SC’s intrinsic fault tolerance and parallelism. Together, these advances mark a transition from randomness-driven SC to engineered, architecture-level bit-stream computing ready for energy-constrained and intermittently powered systems.
Dr. M. Hassan Najafi
Dr. M. Hassan Najafi is an Associate Professor in the Electrical, Computer, and Systems Engineering (ECSE) Department at Case Western Reserve University (CWRU), where he directs the Emerging Computing Technologies (ECT) Laboratory. Prior to joining CWRU, he was an Assistant Professor in the School of Computing and Informatics, University of Louisiana at Lafayette from 2018 to 2024. He received his Ph.D. in Electrical and Electronics Engineering from the University of Minnesota-Twin Cities in 2018. His research is marked by pioneering advances in stochastic computing, unary computing, hyperdimensional computing, and low-power VLSI design. He has published over 110 peer-reviewed articles in leading EDA journals and conferences. He was a recipient of the NSF CAREER Award in 2024 and the winner of the NASA Beyond the Algorithm Challenge Award in 2025. He received the Best Paper Award at GLSVLSI'23 and ICCD’17, the Best Poster Award at GLSVLSI'24, the Best Research Award at the 2019 DAC PhD Forum and the 2018 EDAA Outstanding Dissertation Award. Dr. Najafi has been an editor for the IEEE Journal on Emerging and Selected Topics in Circuits and Systems and a technical committee member and reviewer for more than 100 top IEEE and ACM conferences and journals. Dr. Najafi is an inventor or co-inventor of 23 non-provisional U.S. Patents (10 issued and 13 pending). He is a Senior Member of IEEE and a Senior Member of the U.S. National Academy of Inventors (NAI).
Page updated
Google Sites
Report abuse