Quantized Context Based LIF Neurons for Recurrent Spiking Neural Networks in 45nm
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916227397451776 |
|---|---|
| author | Bezugam, Sai Sukruth Wu, Yihao Yoo, JaeBum Strukov, Dmitri Kim, Bongjin |
| author_facet | Bezugam, Sai Sukruth Wu, Yihao Yoo, JaeBum Strukov, Dmitri Kim, Bongjin |
| contents | In this study, we propose the first hardware implementation of a context-based recurrent spiking neural network (RSNN) emphasizing on integrating dual information streams within the neocortical pyramidal neurons specifically Context- Dependent Leaky Integrate and Fire (CLIF) neuron models, essential element in RSNN. We present a quantized version of the CLIF neuron (qCLIF), developed through a hardware-software codesign approach utilizing the sparse activity of RSNN. Implemented in a 45nm technology node, the qCLIF is compact (900um^2) and achieves a high accuracy of 90% despite 8 bit quantization on DVS gesture classification dataset. Our analysis spans a network configuration from 10 to 200 qCLIF neurons, supporting up to 82k synapses within a 1.86 mm^2 footprint, demonstrating scalability and efficiency |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_18066 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Quantized Context Based LIF Neurons for Recurrent Spiking Neural Networks in 45nm Bezugam, Sai Sukruth Wu, Yihao Yoo, JaeBum Strukov, Dmitri Kim, Bongjin Neural and Evolutionary Computing Artificial Intelligence Hardware Architecture Computer Vision and Pattern Recognition Neurons and Cognition In this study, we propose the first hardware implementation of a context-based recurrent spiking neural network (RSNN) emphasizing on integrating dual information streams within the neocortical pyramidal neurons specifically Context- Dependent Leaky Integrate and Fire (CLIF) neuron models, essential element in RSNN. We present a quantized version of the CLIF neuron (qCLIF), developed through a hardware-software codesign approach utilizing the sparse activity of RSNN. Implemented in a 45nm technology node, the qCLIF is compact (900um^2) and achieves a high accuracy of 90% despite 8 bit quantization on DVS gesture classification dataset. Our analysis spans a network configuration from 10 to 200 qCLIF neurons, supporting up to 82k synapses within a 1.86 mm^2 footprint, demonstrating scalability and efficiency |
| title | Quantized Context Based LIF Neurons for Recurrent Spiking Neural Networks in 45nm |
| topic | Neural and Evolutionary Computing Artificial Intelligence Hardware Architecture Computer Vision and Pattern Recognition Neurons and Cognition |
| url | https://arxiv.org/abs/2404.18066 |