Fair and Comprehensive Benchmarking of Machine Learning Processing Chips

Geoffrey W. Burr; SukHwan Lim; Boris Murmann; Rangharajan Venkatesan; Marian Verhelst

doi:10.1109/MDAT.2021.3063366

IEEE Design and Test

Review

31 May 2022

Fair and Comprehensive Benchmarking of Machine Learning Processing Chips

View publication

Abstract

This article provides an overview of benchmarking strategies at a different set of abstraction levels and related best practices in each, while working to also give an overall view of impact.

Conference paper

Circuit-level benchmarking of access devices for resistive nonvolatile memory arrays

Paper

AI hardware acceleration with analog memory: Microarchitectures for low energy at high speed

Paper

Fully On-Chip MAC at 14 nm Enabled by Accurate Row-Wise Programming of PCM-Based Weights and Parallel Vector-Transport in Duration-Format

Paper

An analog-AI chip for energy-efficient speech recognition and transcription

View all publications

Abstract

Related

Circuit-level benchmarking of access devices for resistive nonvolatile memory arrays

AI hardware acceleration with analog memory: Microarchitectures for low energy at high speed

Fully On-Chip MAC at 14 nm Enabled by Accurate Row-Wise Programming of PCM-Based Weights and Parallel Vector-Transport in Duration-Format

An analog-AI chip for energy-efficient speech recognition and transcription