An Attempt to Devise a Pairwise Ising-Type Maximum Entropy Model Integrated Cost Function for Optimizing SNN Deployment

TL;DR

This paper explores a novel Ising-type maximum entropy model to optimize the deployment of Spiking Neural Networks on neuromorphic hardware, aiming to improve configuration efficiency by capturing network dynamics.

Contribution

It introduces a pairwise Ising model framework to incorporate SNN dynamics into deployment optimization, bridging microscopic neuron interactions with macroscopic network behavior.

Findings

Initial results show underperformance, indicating limitations of the model.

Highlights challenges due to equilibrium assumptions and hardware complexity.

Points out need for further experiments and model refinement.

Abstract

Spiking Neural Networks (SNNs) emulate the spiking behavior of biological neurons and are typically deployed on distributed-memory neuromorphic hardware. The deployment of a SNN usually requires partitioning the network and mapping these partitions onto the hardware's processing units. However, finding optimal deployment configurations is an NP-hard problem, often addressed through optimization algorithms. While some objectives (e.g., memory utilization and chip count) are static, others (e.g., communication latency and energy efficiency) depend on the network's dynamic behavior, necessitating dynamic-aware optimization. To address this, we model SNN dynamics using an Ising-type pairwise interaction framework, bridging microscopic neuron interactions with macroscopic network behavior. We optimize deployment by exploring the parameter and configuration spaces of the Ising model. We evaluate our approach on two SNNs deployed on the sPyNNaker neuromorphic platform. Initial results suggest that the method underperforms, potentially due to the Ising model's equilibrium assumptions and the architectural complexity of real-world neuromorphic hardware, highlighting limitations in its current formulation. Update: The method proposed is with a equilibrium-dynamics SNN assumption, and the original paper does not mention this. The paper needs to be revisited and reuploaded after further experiments.