File Download
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1002/aisy.202100114
- WOS: WOS:000693099800001
- Find via
Supplementary
-
Citations:
- Web of Science: 0
- Appears in Collections:
Article: Energy-Efficient Memristive Euclidean Distance Engine for Brain-Inspired Competitive Learning
Title | Energy-Efficient Memristive Euclidean Distance Engine for Brain-Inspired Competitive Learning |
---|---|
Authors | |
Keywords | Analog computing Competitive learning Euclidean distance engine Memristors |
Issue Date | 2021 |
Publisher | Wiley Open Access. The Journal's web site is located at https://onlinelibrary.wiley.com/journal/26404567 |
Citation | Advanced Intelligent Systems, 2021, v. 3 n. 11, article no. 2100114 How to Cite? |
Abstract | 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. |
Persistent Identifier | http://hdl.handle.net/10722/305804 |
ISSN | 2021 Impact Factor: 7.298 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Zhou, H | - |
dc.contributor.author | Chen, J | - |
dc.contributor.author | Wang, Y | - |
dc.contributor.author | Liu, S | - |
dc.contributor.author | Li, Y | - |
dc.contributor.author | Li, Q | - |
dc.contributor.author | Liu, Q | - |
dc.contributor.author | Wang, Z | - |
dc.contributor.author | He, Y | - |
dc.contributor.author | Xu, H | - |
dc.contributor.author | Miao, X | - |
dc.date.accessioned | 2021-10-20T10:14:33Z | - |
dc.date.available | 2021-10-20T10:14:33Z | - |
dc.date.issued | 2021 | - |
dc.identifier.citation | Advanced Intelligent Systems, 2021, v. 3 n. 11, article no. 2100114 | - |
dc.identifier.issn | 2640-4567 | - |
dc.identifier.uri | http://hdl.handle.net/10722/305804 | - |
dc.description.abstract | 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. | - |
dc.language | eng | - |
dc.publisher | Wiley Open Access. The Journal's web site is located at https://onlinelibrary.wiley.com/journal/26404567 | - |
dc.relation.ispartof | Advanced Intelligent Systems | - |
dc.rights | This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. | - |
dc.subject | Analog computing | - |
dc.subject | Competitive learning | - |
dc.subject | Euclidean distance engine | - |
dc.subject | Memristors | - |
dc.title | Energy-Efficient Memristive Euclidean Distance Engine for Brain-Inspired Competitive Learning | - |
dc.type | Article | - |
dc.identifier.email | Wang, Z: zrwang@eee.hku.hk | - |
dc.identifier.authority | Wang, Z=rp02714 | - |
dc.description.nature | published_or_final_version | - |
dc.identifier.doi | 10.1002/aisy.202100114 | - |
dc.identifier.hkuros | 327761 | - |
dc.identifier.volume | 3 | - |
dc.identifier.issue | 11 | - |
dc.identifier.spage | article no. 2100114 | - |
dc.identifier.epage | article no. 2100114 | - |
dc.identifier.isi | WOS:000693099800001 | - |
dc.publisher.place | Germany | - |