Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system
Abstract Lately, there has been a rapid increase in the use of software-based deep learning neural networks (S-DNN) for the analysis of unstructured data consumption. For implementation of the S-DNN, synapse-device-based hardware DNN (H-DNN) has been proposed as an alternative to typical Von-Neumann...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-12-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-021-02176-4 |
| _version_ | 1818833987941957632 |
|---|---|
| author | Geonhui Han Chuljun Lee Jae-Eun Lee Jongseon Seo Myungjun Kim Yubin Song Young-Ho Seo Daeseok Lee |
| author_facet | Geonhui Han Chuljun Lee Jae-Eun Lee Jongseon Seo Myungjun Kim Yubin Song Young-Ho Seo Daeseok Lee |
| author_sort | Geonhui Han |
| collection | DOAJ |
| description | Abstract Lately, there has been a rapid increase in the use of software-based deep learning neural networks (S-DNN) for the analysis of unstructured data consumption. For implementation of the S-DNN, synapse-device-based hardware DNN (H-DNN) has been proposed as an alternative to typical Von-Neumann structural computing systems. In the H-DNN, various numerical values such as the synaptic weight, activation function, and etc., have to be realized through electrical device or circuit. Among them, the synaptic weight that should have both positive and negative numerical values needs to be implemented in a simpler way. Because the synaptic weight has been expressed by conductance value of the synapse device, it always has a positive value. Therefore, typically, a pair of synapse devices is required to realize the negative weight values, which leads to additional hardware resources such as more devices, higher power consumption, larger area, and increased circuit complexity. Herein, we propose an alternative simpler method to realize the negative weight (named weight shifter) and its hardware implementation. To demonstrate the weight shifter, we investigated its theoretical, numerical, and circuit-related aspects, following which the H-DNN circuit was successfully implemented on a printed circuit board. |
| first_indexed | 2024-12-19T02:27:40Z |
| format | Article |
| id | doaj.art-aec75e692f634a0a90acfdf00a3e5311 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-19T02:27:40Z |
| publishDate | 2021-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-aec75e692f634a0a90acfdf00a3e53112022-12-21T20:39:53ZengNature PortfolioScientific Reports2045-23222021-12-011111710.1038/s41598-021-02176-4Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic systemGeonhui Han0Chuljun Lee1Jae-Eun Lee2Jongseon Seo3Myungjun Kim4Yubin Song5Young-Ho Seo6Daeseok Lee7Department of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityDepartment of Electronic Materials Engineering, Kwangwoon UniversityAbstract Lately, there has been a rapid increase in the use of software-based deep learning neural networks (S-DNN) for the analysis of unstructured data consumption. For implementation of the S-DNN, synapse-device-based hardware DNN (H-DNN) has been proposed as an alternative to typical Von-Neumann structural computing systems. In the H-DNN, various numerical values such as the synaptic weight, activation function, and etc., have to be realized through electrical device or circuit. Among them, the synaptic weight that should have both positive and negative numerical values needs to be implemented in a simpler way. Because the synaptic weight has been expressed by conductance value of the synapse device, it always has a positive value. Therefore, typically, a pair of synapse devices is required to realize the negative weight values, which leads to additional hardware resources such as more devices, higher power consumption, larger area, and increased circuit complexity. Herein, we propose an alternative simpler method to realize the negative weight (named weight shifter) and its hardware implementation. To demonstrate the weight shifter, we investigated its theoretical, numerical, and circuit-related aspects, following which the H-DNN circuit was successfully implemented on a printed circuit board.https://doi.org/10.1038/s41598-021-02176-4 |
| spellingShingle | Geonhui Han Chuljun Lee Jae-Eun Lee Jongseon Seo Myungjun Kim Yubin Song Young-Ho Seo Daeseok Lee Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system Scientific Reports |
| title | Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| title_full | Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| title_fullStr | Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| title_full_unstemmed | Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| title_short | Alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| title_sort | alternative negative weight for simpler hardware implementation of synapse device based neuromorphic system |
| url | https://doi.org/10.1038/s41598-021-02176-4 |
| work_keys_str_mv | AT geonhuihan alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT chuljunlee alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT jaeeunlee alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT jongseonseo alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT myungjunkim alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT yubinsong alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT younghoseo alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem AT daeseoklee alternativenegativeweightforsimplerhardwareimplementationofsynapsedevicebasedneuromorphicsystem |