Skip to main content

Biologically Plausible Learning on Neuromorphic Hardware Architectures

Publication ,  Conference
Wolters, C; Taylor, B; Hanson, E; Yang, X; Schlichtmann, U; Chen, Y
Published in: Midwest Symposium on Circuits and Systems
January 1, 2023

Movement of model parameters from memory to computing elements in deep learning (DL) has led to a growing imbalance known as the memory wall. Neuromorphic computation-in-memory (CIM) is an emerging paradigm that addresses this imbalance by performing computations directly in analog memory. However, sequential backpropagation of error through a network in DL prevents efficient parallelization. A novel method, direct feedback alignment (DFA), resolves layer dependencies by directly passing the error from the output to each layer. This work explores the interrelationship of implementing a bio-plausible learning algorithm like DFA in-situ on neuromorphic CIM hardware, emphasizing energy, area, and latency constraints. Using the DNN+NeuroSim benchmarking framework, we investigate the impact of hardware nonidealities and quantization on algorithm performance, as well as how network topologies and algorithm-level design choices scale latency, energy, and area consumption of a chip. While standard backpropagation learns more accurate models when faced with hardware imperfections, DFA enables significant speedup through parallelization, reducing training time by a factor approaching N for N-layer networks.

Duke Scholars

Published In

Midwest Symposium on Circuits and Systems

DOI

ISSN

1548-3746

Publication Date

January 1, 2023

Start / End Page

733 / 737
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Wolters, C., Taylor, B., Hanson, E., Yang, X., Schlichtmann, U., & Chen, Y. (2023). Biologically Plausible Learning on Neuromorphic Hardware Architectures. In Midwest Symposium on Circuits and Systems (pp. 733–737). https://doi.org/10.1109/MWSCAS57524.2023.10405905
Wolters, C., B. Taylor, E. Hanson, X. Yang, U. Schlichtmann, and Y. Chen. “Biologically Plausible Learning on Neuromorphic Hardware Architectures.” In Midwest Symposium on Circuits and Systems, 733–37, 2023. https://doi.org/10.1109/MWSCAS57524.2023.10405905.
Wolters C, Taylor B, Hanson E, Yang X, Schlichtmann U, Chen Y. Biologically Plausible Learning on Neuromorphic Hardware Architectures. In: Midwest Symposium on Circuits and Systems. 2023. p. 733–7.
Wolters, C., et al. “Biologically Plausible Learning on Neuromorphic Hardware Architectures.” Midwest Symposium on Circuits and Systems, 2023, pp. 733–37. Scopus, doi:10.1109/MWSCAS57524.2023.10405905.
Wolters C, Taylor B, Hanson E, Yang X, Schlichtmann U, Chen Y. Biologically Plausible Learning on Neuromorphic Hardware Architectures. Midwest Symposium on Circuits and Systems. 2023. p. 733–737.

Published In

Midwest Symposium on Circuits and Systems

DOI

ISSN

1548-3746

Publication Date

January 1, 2023

Start / End Page

733 / 737