Public Lecture #1

"Cross-Layer Design Space Exploration for In-Memory Computing Based Accelerators"

Prof. Sharon HU (University of Notre Dame)

Abstract

In-memory computing (IMC) has the potential to significantly ease the data-transfer bottleneck facing many data-intensive applications such as many machine learning, bioinformatics, and secure information processing. However, due to the myriad of design options for an IMC-based solution, harnessing the benefits of the IMC paradigm requires cross-layer efforts spanning from devices and circuits to architectures and applications. This talk presents the design, evaluation, and use of modeling tools to support such cross-layer efforts, i.e., to accurately and rapidly assess the impact of technology-driven IMC architectures that can offer meaningful benefits to applications of the most relevance. The talk will be framed in the context of tools and methodologies that are specifically developed for associative memory (a.k.a. content addressable memory) based IMC-type accelerators and ferroelectric FETs (FeFETs). IMC-based accelerators for several popular machine learning and other applications will be analyzed end-to-end to demonstrate the value contributed by each design layer, which can serve as guides for future research efforts.

Back