Core interface optimization for multi-core neuromorphic processors

TL;DR

This paper introduces an optimized core interface for multi-core neuromorphic processors, significantly reducing latency and energy consumption while increasing throughput, enabling scalable and efficient large-scale SNN implementations.

Contribution

It proposes a hierarchical arbiter tree for arbitration and an asynchronous CAM with CSCD for routing memory, enhancing performance and efficiency in neuromorphic core interfaces.

Findings

Latency reduced by over 70% in sparse-event mode

Routing memory energy consumption decreased by approximately 46%

Throughput increased by 40% with minimal area increase

Abstract

Hardware implementations of Spiking Neural Networks (SNNs) represent a promising approach to edge-computing for applications that require low-power and low-latency, and which cannot resort to external cloud-based computing services. However, most solutions proposed so far either support only relatively small networks, or take up significant hardware resources, to implement large networks. To realize large-scale and scalable SNNs it is necessary to develop an efficient asynchronous communication and routing fabric that enables the design of multi-core architectures. In particular the core interface that manages inter-core spike communication is a crucial component as it represents the bottleneck of Power-Performance-Area (PPA) especially for the arbitration architecture and the routing memory. In this paper we present an arbitration mechanism with the corresponding asynchronous encoding pipeline circuits, based on hierarchical arbiter trees. The proposed scheme reduces the latency by more than 70% in sparse-event mode, compared to the state-of-the-art arbitration architectures, with lower area cost. The routing memory makes use of asynchronous Content Addressable Memory (CAM) with Current Sensing Completion Detection (CSCD), which saves approximately 46% energy, and achieves a 40% increase in throughput against conventional asynchronous CAM using configurable delay lines, at the cost of only a slight increase in area. In addition as it radically reduces the core interface resources in multi-core neuromorphic processors, the arbitration architecture and CAM architecture we propose can be also applied to a wide range of general asynchronous circuits and systems.