SENMAP: Multi-objective data-flow mapping and synthesis for hybrid scalable neuromorphic systems

2025-06-03

AI Summary

Efficiently deploying large-scale spiking neural networks (SNNs) and artificial neural networks (ANNs) on scalable neuromorphic hardware—specifically the SENEC A chip—remains challenging due to the need for joint optimization of energy efficiency, throughput, area, and accuracy. Method: This paper introduces SENMap, a multi-objective mapping and synthesis tool that integrates event-rate modeling, pre-trained model adaptation, dataflow graph decomposition, and multi-objective heuristic search—constituting the first unified mapping framework for SNN/ANN deployment on SENEC A. Contribution/Results: Open-sourced and compatible with the SENSIM simulator, SENMap enables pre-silicon exploration of heterogeneous architectures. Under SENSIM’s asynchronous timing mode, it achieves a 40% reduction in energy consumption over baseline approaches while significantly improving mapping efficiency and scalability for large SNNs. The methodology provides a reusable, co-design foundation for brain-inspired computing chips, bridging algorithmic innovation and neuromorphic hardware constraints.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Hardware-aware ML
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

This paper introduces SENMap, a mapping and synthesis tool for scalable, energy-efficient neuromorphic computing architecture frameworks. SENECA is a flexible architectural design optimized for executing edge AI SNN/ANN inference applications efficiently. To speed up the silicon tape-out and chip design for SENECA, an accurate emulator, SENSIM, was designed. While SENSIM supports direct mapping of SNNs on neuromorphic architectures, as the SNN and ANNs grow in size, achieving optimal mapping for objectives like energy, throughput, area, and accuracy becomes challenging. This paper introduces SENMap, flexible mapping software for efficiently mapping large SNN and ANN applications onto adaptable architectures. SENMap considers architectural, pretrained SNN and ANN realistic examples, and event rate-based parameters and is open-sourced along with SENSIM to aid flexible neuromorphic chip design before fabrication. Experimental results show SENMap enables 40 percent energy improvements for a baseline SENSIM operating in timestep asynchronous mode of operation. SENMap is designed in such a way that it facilitates mapping large spiking neural networks for future modifications as well.

Problem

• Optimizes mapping of large SNNs/ANNs for energy and performance
• Addresses challenges in scalable neuromorphic system design
• Enables flexible architecture adaptation for edge AI applications

Innovation

• Flexible mapping software for large SNN/ANN
• Accurate emulator SENSIM for chip design
• Energy-efficient neuromorphic computing architecture