Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator
Abstract
The training process for spiking neural networks can be very computationally intensive. Approaches such as evolutionary algorithms may require evaluating thousands or millions of candidate solutions. In this work, we propose using neuromorphic cores implemented on a Xilinx Zynq system on chip to accelerate and improve the energy efficiency of the evaluation step of an evolutionary training approach. We demonstrate this can significantly reduce the required energy to evolve a network with some cases showing greater than 10 times improvement as compared to a CPU-only system.
- Authors:
-
- ORNL
- Publication Date:
- Research Org.:
- Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)
- OSTI Identifier:
- 1827021
- DOE Contract Number:
- AC05-00OR22725
- Resource Type:
- Conference
- Resource Relation:
- Conference: International Conference on Neuromorphic Systems (ICONS) 2021 - Knoxville, Tennessee, United States of America - 7/27/2021 5:00:00 AM-7/29/2021 5:00:00 AM
- Country of Publication:
- United States
- Language:
- English
Citation Formats
Mitchell, Parker, and Schuman, Catherine. Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator. United States: N. p., 2021.
Web. doi:10.1145/3477145.3477150.
Mitchell, Parker, & Schuman, Catherine. Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator. United States. https://doi.org/10.1145/3477145.3477150
Mitchell, Parker, and Schuman, Catherine. 2021.
"Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator". United States. https://doi.org/10.1145/3477145.3477150. https://www.osti.gov/servlets/purl/1827021.
@article{osti_1827021,
title = {Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator},
author = {Mitchell, Parker and Schuman, Catherine},
abstractNote = {The training process for spiking neural networks can be very computationally intensive. Approaches such as evolutionary algorithms may require evaluating thousands or millions of candidate solutions. In this work, we propose using neuromorphic cores implemented on a Xilinx Zynq system on chip to accelerate and improve the energy efficiency of the evaluation step of an evolutionary training approach. We demonstrate this can significantly reduce the required energy to evolve a network with some cases showing greater than 10 times improvement as compared to a CPU-only system.},
doi = {10.1145/3477145.3477150},
url = {https://www.osti.gov/biblio/1827021},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Thu Jul 01 00:00:00 EDT 2021},
month = {Thu Jul 01 00:00:00 EDT 2021}
}
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.
Save to My Library
You must Sign In or Create an Account in order to save documents to your library.