Spiking Reservoir Computing Based on Stochastic Diffusive Memristors
Abstract Reservoir computing (RC), a type of recurrent neural network, is particularly well‐suited for hardware implementation in edge computing. It is shown that RC hardware based on dynamic memristors potentially offers much lower power consumption and reduced computation times than digital electr...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley-VCH
2025-03-01
|
| Series: | Advanced Electronic Materials |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/aelm.202400469 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850032519579172864 |
|---|---|
| author | Zelin Ma Jun Ge Shusheng Pan |
| author_facet | Zelin Ma Jun Ge Shusheng Pan |
| author_sort | Zelin Ma |
| collection | DOAJ |
| description | Abstract Reservoir computing (RC), a type of recurrent neural network, is particularly well‐suited for hardware implementation in edge computing. It is shown that RC hardware based on dynamic memristors potentially offers much lower power consumption and reduced computation times than digital electronics. However, challenges such as stochasticity and read noise in these devices can impair its performance. Furthermore, the external analog‐to‐digital (ADC) readout circuits may require substantial area and energy. In this work, it is experimentally demonstrated that a population of stochastic diffusive Ag:SiOx memristors can effectively construct a spiking reservoir computing system. This system demonstrates remarkable resilience to read noise and delivers exceptional performance across a range of computational tasks, achieving a 98% accuracy in waveform classification and a normalized root mean square error (NRMSE) of 0.154 in time‐series prediction. Further simulations reveal that a certain degree of device stochasticity actually enhances system performance. Without using ADC converters, a hybrid memristor‐CMOS spiking RC system is designed that demonstrates significantly lower power consumption compared to fully digital systems. |
| format | Article |
| id | doaj-art-1229d4e421124627947b01a098e4100a |
| institution | DOAJ |
| issn | 2199-160X |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Wiley-VCH |
| record_format | Article |
| series | Advanced Electronic Materials |
| spelling | doaj-art-1229d4e421124627947b01a098e4100a2025-08-20T02:58:37ZengWiley-VCHAdvanced Electronic Materials2199-160X2025-03-01113n/an/a10.1002/aelm.202400469Spiking Reservoir Computing Based on Stochastic Diffusive MemristorsZelin Ma0Jun Ge1Shusheng Pan2Key Lab of Si‐based Information Materials & Devices and Integrated Circuits Design Department of Education of Guangdong Province Guangzhou Higher Education Mega Center Panyu District Guangzhou 510006 ChinaKey Lab of Si‐based Information Materials & Devices and Integrated Circuits Design Department of Education of Guangdong Province Guangzhou Higher Education Mega Center Panyu District Guangzhou 510006 ChinaKey Lab of Si‐based Information Materials & Devices and Integrated Circuits Design Department of Education of Guangdong Province Guangzhou Higher Education Mega Center Panyu District Guangzhou 510006 ChinaAbstract Reservoir computing (RC), a type of recurrent neural network, is particularly well‐suited for hardware implementation in edge computing. It is shown that RC hardware based on dynamic memristors potentially offers much lower power consumption and reduced computation times than digital electronics. However, challenges such as stochasticity and read noise in these devices can impair its performance. Furthermore, the external analog‐to‐digital (ADC) readout circuits may require substantial area and energy. In this work, it is experimentally demonstrated that a population of stochastic diffusive Ag:SiOx memristors can effectively construct a spiking reservoir computing system. This system demonstrates remarkable resilience to read noise and delivers exceptional performance across a range of computational tasks, achieving a 98% accuracy in waveform classification and a normalized root mean square error (NRMSE) of 0.154 in time‐series prediction. Further simulations reveal that a certain degree of device stochasticity actually enhances system performance. Without using ADC converters, a hybrid memristor‐CMOS spiking RC system is designed that demonstrates significantly lower power consumption compared to fully digital systems.https://doi.org/10.1002/aelm.202400469memristorsphysical reservoir computingpopulationsstochastic electronics |
| spellingShingle | Zelin Ma Jun Ge Shusheng Pan Spiking Reservoir Computing Based on Stochastic Diffusive Memristors Advanced Electronic Materials memristors physical reservoir computing populations stochastic electronics |
| title | Spiking Reservoir Computing Based on Stochastic Diffusive Memristors |
| title_full | Spiking Reservoir Computing Based on Stochastic Diffusive Memristors |
| title_fullStr | Spiking Reservoir Computing Based on Stochastic Diffusive Memristors |
| title_full_unstemmed | Spiking Reservoir Computing Based on Stochastic Diffusive Memristors |
| title_short | Spiking Reservoir Computing Based on Stochastic Diffusive Memristors |
| title_sort | spiking reservoir computing based on stochastic diffusive memristors |
| topic | memristors physical reservoir computing populations stochastic electronics |
| url | https://doi.org/10.1002/aelm.202400469 |
| work_keys_str_mv | AT zelinma spikingreservoircomputingbasedonstochasticdiffusivememristors AT junge spikingreservoircomputingbasedonstochasticdiffusivememristors AT shushengpan spikingreservoircomputingbasedonstochasticdiffusivememristors |