Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Brehove, Matthew, Tumpa, Sadia Anjum, Kyubwa, Espoir, Menon, Naresh, Narayanan, Vijaykrishnan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2505.06417
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912368708026368
author Brehove, Matthew
Tumpa, Sadia Anjum
Kyubwa, Espoir
Menon, Naresh
Narayanan, Vijaykrishnan
author_facet Brehove, Matthew
Tumpa, Sadia Anjum
Kyubwa, Espoir
Menon, Naresh
Narayanan, Vijaykrishnan
contents Neuromorphic computing aims to improve the efficiency of artificial neural networks by taking inspiration from biological neurons and leveraging temporal sparsity, spatial sparsity, and compute near/in memory. Although these approaches have shown efficiency gains, training these spiking neural networks (SNN) remains difficult. The original attempts at converting trained conventional analog neural networks (ANN) to SNNs used the rate of binary spikes to represent neuron activations. This required many simulation time steps per inference, which degraded efficiency. Intel's Loihi 2 is a neuromorphic platform that supports graded spikes which can be used to represent changes in neuron activation. In this work, we use Loihi 2's graded spikes to develop a method for converting ANN networks to spiking networks, which take advantage of temporal and spatial sparsity. We evaluated the performance of this network on Loihi 2 and compared it to NVIDIA's Jetson Xavier edge AI platform.
format Preprint
id arxiv_https___arxiv_org_abs_2505_06417
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Sigma-Delta Neural Network Conversion on Loihi 2
Brehove, Matthew
Tumpa, Sadia Anjum
Kyubwa, Espoir
Menon, Naresh
Narayanan, Vijaykrishnan
Neural and Evolutionary Computing
Neuromorphic computing aims to improve the efficiency of artificial neural networks by taking inspiration from biological neurons and leveraging temporal sparsity, spatial sparsity, and compute near/in memory. Although these approaches have shown efficiency gains, training these spiking neural networks (SNN) remains difficult. The original attempts at converting trained conventional analog neural networks (ANN) to SNNs used the rate of binary spikes to represent neuron activations. This required many simulation time steps per inference, which degraded efficiency. Intel's Loihi 2 is a neuromorphic platform that supports graded spikes which can be used to represent changes in neuron activation. In this work, we use Loihi 2's graded spikes to develop a method for converting ANN networks to spiking networks, which take advantage of temporal and spatial sparsity. We evaluated the performance of this network on Loihi 2 and compared it to NVIDIA's Jetson Xavier edge AI platform.
title Sigma-Delta Neural Network Conversion on Loihi 2
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2505.06417