Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning
Inspired by competitive rules of the nature, competitive learning contributes to the specialization of the human brain and the general creativity of mankind. However, the construction of hardware competitive learning neural network still faces great challenges due to the lack of an accurate distance...
Guardado en:
Autores principales: | , , , , , , , , , , |
---|---|
Formato: | article |
Lenguaje: | EN |
Publicado: |
Wiley
2021
|
Materias: | |
Acceso en línea: | https://doaj.org/article/9b241c76c298406f8ec126d570d6e327 |
Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
id |
oai:doaj.org-article:9b241c76c298406f8ec126d570d6e327 |
---|---|
record_format |
dspace |
spelling |
oai:doaj.org-article:9b241c76c298406f8ec126d570d6e3272021-11-23T07:58:48ZEnergy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning2640-456710.1002/aisy.202100114https://doaj.org/article/9b241c76c298406f8ec126d570d6e3272021-11-01T00:00:00Zhttps://doi.org/10.1002/aisy.202100114https://doaj.org/toc/2640-4567Inspired by competitive rules of the nature, competitive learning contributes to the specialization of the human brain and the general creativity of mankind. However, the construction of hardware competitive learning neural network still faces great challenges due to the lack of an accurate distance computation method and a self‐adaptive in situ training scheme. Herein, a fully memristive Euclidean distance (ED) engine based on analog multiply‐accumulate operation in a 32 × 32 TiN/TaO x /HfO x /TiN 1T1R array is demonstrated. The dual‐layer devices perform multilevel modulation under the target‐aware programming method with excellent read linearity in a dynamic range of 10–100 μS. The ED calculation is verified experimentally on a test board with an O(1) temporal complexity. Furthermore, in situ training and offline inference schemes for competitive learning, based on the ED engine, are developed and the simulated results show comparable success rates with those obtained by the CPU‐based software. Compared with a state‐of‐the‐art RTX6000 GPU (0.5 TOPS W−1), the energy efficiency of competitive learning models on ED engines can yield 100× improvements by utilizing optimized memristive devices.Houji ZhouJia ChenYinan WangSen LiuYi LiQingjiang LiQi LiuZhongrui WangYuhui HeHui XuXiangshui MiaoWileyarticleanalog computingcompetitive learningEuclidean distance enginememristorsComputer engineering. Computer hardwareTK7885-7895Control engineering systems. Automatic machinery (General)TJ212-225ENAdvanced Intelligent Systems, Vol 3, Iss 11, Pp n/a-n/a (2021) |
institution |
DOAJ |
collection |
DOAJ |
language |
EN |
topic |
analog computing competitive learning Euclidean distance engine memristors Computer engineering. Computer hardware TK7885-7895 Control engineering systems. Automatic machinery (General) TJ212-225 |
spellingShingle |
analog computing competitive learning Euclidean distance engine memristors Computer engineering. Computer hardware TK7885-7895 Control engineering systems. Automatic machinery (General) TJ212-225 Houji Zhou Jia Chen Yinan Wang Sen Liu Yi Li Qingjiang Li Qi Liu Zhongrui Wang Yuhui He Hui Xu Xiangshui Miao Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
description |
Inspired by competitive rules of the nature, competitive learning contributes to the specialization of the human brain and the general creativity of mankind. However, the construction of hardware competitive learning neural network still faces great challenges due to the lack of an accurate distance computation method and a self‐adaptive in situ training scheme. Herein, a fully memristive Euclidean distance (ED) engine based on analog multiply‐accumulate operation in a 32 × 32 TiN/TaO x /HfO x /TiN 1T1R array is demonstrated. The dual‐layer devices perform multilevel modulation under the target‐aware programming method with excellent read linearity in a dynamic range of 10–100 μS. The ED calculation is verified experimentally on a test board with an O(1) temporal complexity. Furthermore, in situ training and offline inference schemes for competitive learning, based on the ED engine, are developed and the simulated results show comparable success rates with those obtained by the CPU‐based software. Compared with a state‐of‐the‐art RTX6000 GPU (0.5 TOPS W−1), the energy efficiency of competitive learning models on ED engines can yield 100× improvements by utilizing optimized memristive devices. |
format |
article |
author |
Houji Zhou Jia Chen Yinan Wang Sen Liu Yi Li Qingjiang Li Qi Liu Zhongrui Wang Yuhui He Hui Xu Xiangshui Miao |
author_facet |
Houji Zhou Jia Chen Yinan Wang Sen Liu Yi Li Qingjiang Li Qi Liu Zhongrui Wang Yuhui He Hui Xu Xiangshui Miao |
author_sort |
Houji Zhou |
title |
Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
title_short |
Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
title_full |
Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
title_fullStr |
Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
title_full_unstemmed |
Energy‐Efficient Memristive Euclidean Distance Engine for Brain‐Inspired Competitive Learning |
title_sort |
energy‐efficient memristive euclidean distance engine for brain‐inspired competitive learning |
publisher |
Wiley |
publishDate |
2021 |
url |
https://doaj.org/article/9b241c76c298406f8ec126d570d6e327 |
work_keys_str_mv |
AT houjizhou energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT jiachen energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT yinanwang energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT senliu energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT yili energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT qingjiangli energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT qiliu energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT zhongruiwang energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT yuhuihe energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT huixu energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning AT xiangshuimiao energyefficientmemristiveeuclideandistanceengineforbraininspiredcompetitivelearning |
_version_ |
1718416842616209408 |