DFSynthesizer: Dataflow-based Synthesis of Spiking Neural Networks to Neuromorphic Hardware

TL;DR

DFSynthesizer is an end-to-end framework that synthesizes spiking neural networks onto neuromorphic hardware, optimizing resource utilization and performance guarantees through workload analysis, clustering, and scheduling.

Contribution

It introduces a novel, comprehensive approach for mapping SNNs onto neuromorphic hardware, addressing resource and latency constraints effectively.

Findings

Provides tighter performance guarantees than existing methods

Successfully maps 10 machine-learning programs onto neuromorphic hardware

Improves hardware utilization and throughput

Abstract

Spiking Neural Networks (SNN) are an emerging computation model, which uses event-driven activation and bio-inspired learning algorithms. SNN-based machine-learning programs are typically executed on tile- based neuromorphic hardware platforms, where each tile consists of a computation unit called crossbar, which maps neurons and synapses of the program. However, synthesizing such programs on an off-the-shelf neuromorphic hardware is challenging. This is because of the inherent resource and latency limitations of the hardware, which impact both model performance, e.g., accuracy, and hardware performance, e.g., throughput. We propose DFSynthesizer, an end-to-end framework for synthesizing SNN-based machine learning programs to neuromorphic hardware. The proposed framework works in four steps. First, it analyzes a machine-learning program and generates SNN workload using representative data. Second, it partitions the SNN workload and generates clusters that fit on crossbars of the target neuromorphic hardware. Third, it exploits the rich semantics of Synchronous Dataflow Graph (SDFG) to represent a clustered SNN program, allowing for performance analysis in terms of key hardware constraints such as number of crossbars, dimension of each crossbar, buffer space on tiles, and tile communication bandwidth. Finally, it uses a novel scheduling algorithm to execute clusters on crossbars of the hardware, guaranteeing hardware performance. We evaluate DFSynthesizer with 10 commonly used machine-learning programs. Our results demonstrate that DFSynthesizer provides much tighter performance guarantee compared to current mapping approaches.