An On-Chip Trainable Neuron Circuit for SFQ-Based Spiking Neural Networks

被引：2

作者：

Ucpinar, Beyza Zeynep ^{[1
]}

Karamuftuoglu, Mustafa Altay ^{[1
]}

Razmkhah, Sasan ^{[1
]}

Pedram, Massoud ^{[1
]}

机构：

[1] Univ Southern Calif, Ming Hsieh Dept Elect & Comp Engn, Los Angeles, CA 90007 USA

来源：

IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY | 2024年 / 34卷 / 03期

基金：

美国国家科学基金会;

关键词：

Neurons; Biological neural networks; Training; Superconductivity; System-on-chip; Clocks; Simulation; Adjustable neuron; on-chip training; SFQ; spiking neural network;

D O I：

10.1109/TASC.2024.3359164

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

We present an on-chip trainable neuron circuit. Our proposed circuit aims at bio-inspired spike-based time-dependent data computation for training spiking neural networks (SNN). The thresholds of neurons can be increased or decreased depending on the desired application-specific spike generation rate. This mechanism is scalable and provides us with a flexible circuit structure design. We simulated the trainable neuron structure under different operating scenarios with thermal noise included. The circuits are designed and optimized for the MIT LL SFQ5ee fabrication process. For a 16-input neuron with four different threshold values, all of the circuit parameter margins are above 20% (+/- 10%) with a 3G sample per second throughput.

引用

页码：1 / 6

页数：6

共 36 条

[1]

[Anonymous], 1989, Rev. A, V40, P1036, DOI [10.1103/PhysRevA.40.1036[18]X, DOI 10.1103/PHYSREVA.40.1036[18]X]

[2]

Bozbey A, 2020, Arxiv, DOI arXiv:1812.10354

[3] An adaptive threshold mechanism for accurate and efficient deep spiking convolutional neural networks [J].

Chen, Yunhua ;

Mai, Yingchao ;

Feng, Ren ;

Xiao, Jinsheng .

NEUROCOMPUTING, 2022, 469 :189-197

[4] Biologically Realistic Behaviors From a Superconducting Neuron Model [J].

Crotty, Patrick ;

Segall, Kenneth ;

Schult, Daniel .

IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, 2023, 33 (04)

[5] Unsupervised learning of digit recognition using spike-timing-dependent plasticity [J].

Diehl, Peter U. ;

Cook, Matthew .

FRONTIERS IN COMPUTATIONAL NEUROSCIENCE, 2015, 9

[6] Training Spiking Neural Networks Using Lessons From Deep Learning [J].

Eshraghian, Jason K. ;

Ward, Max ;

Neftci, Emre O. ;

Wang, Xinxin ;

Lenz, Gregor ;

Dwivedi, Girish ;

Bennamoun, Mohammed ;

Jeong, Doo Seok ;

Lu, Wei D. .

PROCEEDINGS OF THE IEEE, 2023, 111 (09) :1016-1054

[7]

Fang E. S., 1989, P INT SUP EL C

[8] Results From the ColdFlux Superconductor Integrated Circuit Design Tool Project [J].

Fourie, Coenrad J. ;

Jackman, Kyle ;

Delport, Johannes ;

Schindler, Lieze ;

Hall, Tessa ;

Febvre, Pascal ;

Iwanikow, Lucas ;

Chen, Olivia ;

Ayala, Christopher L. ;

Yoshikawa, Nobuyuki ;

Law, Mark ;

Weingartner, Thomas A. ;

Wang, Yanzhi ;

Beerel, Peter ;

Gupta, Sandeep ;

Zha, Haipeng ;

Razmkhah, Sasan ;

Karamuftuoglu, Mustafa Altay ;

Fayyazi, Arash ;

Li, Mingye ;

Annavaram, Murali ;

Nazarian, Shahin ;

Pedram, Massoud .

IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, 2023, 33 (08)

[9] Large-scale neuromorphic computing systems [J].

Furber, Steve .

JOURNAL OF NEURAL ENGINEERING, 2016, 13 (05)

[10] Spiking Optical Flow for Event-Based Sensors Using IBM's TrueNorth Neurosynaptic System [J].

Haessig, Germain ;

Cassidy, Andrew ;

Alvarez, Rodrigo ;

Benosman, Ryad ;

Orchard, Garrick .

IEEE TRANSACTIONS ON BIOMEDICAL CIRCUITS AND SYSTEMS, 2018, 12 (04) :860-870

← 1 2 3 4 →