Hardware-aware training for large-scale and diverse deep learning inference workloads using in-memory computing-based accelerators

Malte J. Rasch (), Charles Mackin, Manuel Gallo, An Chen, Andrea Fasoli, Frédéric Odermatt, Ning Li, S. R. Nandakumar, Pritish Narayanan, Hsinyu Tsai, Geoffrey W. Burr, Abu Sebastian and Vijay Narayanan
Additional contact information
Malte J. Rasch: TJ Watson Research Center
Charles Mackin: IBM Research Almaden
Manuel Gallo: IBM Research Europe
An Chen: IBM Research Almaden
Andrea Fasoli: IBM Research Almaden
Frédéric Odermatt: IBM Research Europe
Ning Li: TJ Watson Research Center
S. R. Nandakumar: IBM Research Europe
Pritish Narayanan: IBM Research Almaden
Hsinyu Tsai: IBM Research Almaden
Geoffrey W. Burr: IBM Research Almaden
Abu Sebastian: IBM Research Europe
Vijay Narayanan: TJ Watson Research Center

Nature Communications, 2023, vol. 14, issue 1, 1-18

Abstract: Abstract Analog in-memory computing—a promising approach for energy-efficient acceleration of deep learning workloads—computes matrix-vector multiplications but only approximately, due to nonidealities that often are non-deterministic or nonlinear. This can adversely impact the achievable inference accuracy. Here, we develop an hardware-aware retraining approach to systematically examine the accuracy of analog in-memory computing across multiple network topologies, and investigate sensitivity and robustness to a broad set of nonidealities. By introducing a realistic crossbar model, we improve significantly on earlier retraining approaches. We show that many larger-scale deep neural networks—including convnets, recurrent networks, and transformers—can in fact be successfully retrained to show iso-accuracy with the floating point implementation. Our results further suggest that nonidealities that add noise to the inputs or outputs, not the weights, have the largest impact on accuracy, and that recurrent networks are particularly robust to all nonidealities.

Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.nature.com/articles/s41467-023-40770-4 Abstract (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:natcom:v:14:y:2023:i:1:d:10.1038_s41467-023-40770-4

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-023-40770-4

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().