SSCS Open Journal Webinar: MR-PIPA: An Integrated Multilevel RRAM (HfOx)-Based Processing-In-Pixel Accelerator, Presented by Shaahin Angizi & Arman Roohi

Lecture

MR-PIPA: An Integrated Multilevel RRAM (HfOx)-Based Processing-In-Pixel Accelerator

Date

2023-08-03

Time

10:00 AM ET

Location

Webinar - Online

Contact

Danielle Marinese – d.marinese@ieee.org

Sponsorship

Sponsor

Presenter

Shaahin Angizi & Arman Roohi

Description

Abstract: Internet of Things (IoT) devices are projected to attain a $1100B market by 2025, with a web of interconnection projected to comprise approximately 75+ billion IoT devices. The large number of IoTs consist of sensory systems that enable massive data collection from the environment and people. However, considerable portions of the captured sensory data are redundant and unstructured. Data conversion of such large raw data, storing in volatile memories, transmission, and computation in on-/off-chip processors, impose high energy consumption, latency, and a memory bottleneck at the edge. Motivated by the aforementioned concerns, this work paves the way to realize a processing-in-pixel accelerator based on a multi-level HfOx RRAM as a flexible, energy-efficient, and high-performance solution for real-time and smart image processing at edge devices. The proposed design intrinsically implements and supports a coarse-grained convolution operation in low-bit-width neural networks leveraging a novel compute-pixel with non-volatile weight storage at the sensor side. Our evaluations show that such a design can remarkably reduce the power consumption of data conversion and transmission to an off-chip processor maintaining accuracy compared with the recent in-sensor computing designs. Our proposed design, namely MR-PIPA, achieves a frame rate of 1000 and efficiency of ~1.89 TOp/s/W, while it substantially reduces data conversion and transmission energy by ~84% compared to a baseline at the cost of minor accuracy degradation.

 

Bio: Shaahin Angizi (IEEE Senior Member) is currently an Assistant Professor in the Department of Electrical and Computer Engineering, New Jersey Institute of Technology (NJIT), Newark, NJ, USA, and the director of the Advanced Circuit-to-Architecture Design Laboratory. He completed his doctoral studies in Electrical Engineering at the School of Electrical, Computer and Energy Engineering, Arizona State University (ASU), Tempe, AZ in 2021. His primary research interests include ultra-low-power in-memory computing based on volatile & non-volatile memories, in-sensor computing for IoT, brain-inspired (neuromorphic) computing, and accelerator design for deep neural networks and bioinformatics. He has authored and co-authored +90 research papers in top-ranked journals and EDA conferences. He is the recipient of the Best Ph.D. Research Award (1st-place) of Ph.D. Forum at IEEE/ACM DAC in 2018, two Best Paper Awards of IEEE ISVLSI in 2017 and 2018, and Best Paper Award of ACM GLSVLSI in 2019. For more information, please see http://shaahinangizi.com.

Arman Roohi (IEEE Senior Member) is currently an assistant professor with the School Computing, University of Nebraska-Lincoln, USA. Before joining UNL in 2020, he was a postdoctoral research fellow at the University of Texas at Austin. He received Ph.D. degree in Computer Engineering at the University of Central Florida, Orlando, FL, USA, in 2019. His research interests span the areas of design of cross-layer co-design for implementing complex machine learning tasks secure computation, including hardware security, and the security of artificial intelligence, reconfigurable and adaptive computer architectures, and beyond CMOS computing. He has completed over 50 publications on these topics, including best paper recognition, book chapters, and STEM curricular development. He received Ph.D. Forum at DAC 2018 Scholarship, Frank Hubbard Engineering Endowed Scholarship in 2018, best paper recognitions in IEEE Transactions on Emerging Topics in Computing in 2019, and paper of the month at IEEE Transactions on Computers in 2017. For more information, please see https://armanroohi.com/.