Abstract | ||
---|---|---|
Neuromorphic computing has emerged as a promising avenue towards building the next generation of intelligent computing systems. It has been proposed that memristive devices, which exhibit history-dependent conductivity modulation, could efficiently represent the synaptic weights in artificial neural networks. However, precise modulation of the device conductance over a wide dynamic range, necessary to maintain high network accuracy, is proving to be challenging. To address this, we present a multi-memristive synaptic architecture with an efficient global counter-based arbitration scheme. We focus on phase change memory devices, develop a comprehensive model and demonstrate via simulations the effectiveness of the concept for both spiking and non-spiking neural networks. Moreover, we present experimental results involving over a million phase change memory devices for unsupervised learning of temporal correlations using a spiking neural network. The work presents a significant step towards the realization of large-scale and energy-efficient neuromorphic computing systems. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1038/s41467-018-04933-y | NATURE COMMUNICATIONS |
Field | DocType | Volume |
Wide dynamic range,Phase-change memory,Computer architecture,Biology,Neuromorphic engineering,Modulation,Unsupervised learning,Artificial neural network,Spiking neural network,Genetics,Network model | Journal | 9 |
Issue | ISSN | Citations |
1 | 2041-1723 | 9 |
PageRank | References | Authors |
0.82 | 17 | 10 |
Name | Order | Citations | PageRank |
---|---|---|---|
Irem Boybat | 1 | 34 | 5.41 |
Manuel Le Gallo | 2 | 47 | 9.73 |
S. R. Nandakumar | 3 | 10 | 3.67 |
Timoleon Moraitis | 4 | 10 | 1.52 |
Thomas P. Parnell | 5 | 43 | 4.34 |
Tomas Tuma | 6 | 41 | 4.61 |
Bipin Rajendran | 7 | 47 | 12.17 |
Yusuf Leblebici | 8 | 771 | 119.09 |
Sebastian, A. | 9 | 267 | 44.35 |
Evangelos Eleftheriou | 10 | 1590 | 118.20 |