重新评估脉冲神经网络推理能效，发现其仅在特定条件下更省电。
Reconsidering the Energy Efficiency of Spiking Neural Networks Inference from Analytical Perspectives
- 通过匹配容量的量化神经网络作对比，公平评估能效
- 当脉冲率低于5.7%时，脉冲网络在典型硬件下更节能
- 揭示了真实能效优势的适用条件，指导高效系统设计
脉冲神经网络（SNNs）因事件驱动的脉冲计算，理论上比传统量化神经网络（QNNs）更具能效优势。然而，现有能效评估常过于简化，忽略数据传输和内存访问等关键开销，导致结论失真。本文提出严谨重评：将时间编码的SNN映射为位数匹配的QNN（$ ceil ext{log}_2(T+1) ceil$位），确保两者表征能力与硬件需求相当，实现公平比较。构建包含核心计算与数据移动的解析能效模型，系统分析多种参数组合，包括网络特性（时间窗大小 $T$、脉冲率 $s_r$、稀疏性、模型规模、权重量化位数）与硬件特性（存储系统、片上网络）。结果表明，在典型类脑硬件条件下，当 $T=5$ 且平均脉冲率 $s_r < 5.7 ext{ extperthousand}$ 时，SNN 才真正优于等效 QNN。该研究为设计真正高效的神经网络方案提供关键指引。
原文摘要 · Abstract (English)
Spiking Neural Networks (SNNs) promise higher energy efficiency over conventional Quantized Artificial Neural Networks (QNNs) due to their event-driven, spike-based computation. However, prevailing energy evaluations often oversimplify, focusing on computational aspects while neglecting critical overheads like comprehensive data movements and memory accesses. Such simplifications can lead to misleading conclusions regarding the true energy benefits of SNNs. This paper presents a rigorous re-evaluation. We establish a fair baseline by mapping rate-encoded SNNs with $T$ timesteps to capacity-matched QNNs with $\lceil \log_2(T+1) \rceil$ bits. This ensures both models have comparable representational capacities, as well as similar hardware requirements, enabling meaningful energy comparisons. We introduce a detailed analytical energy model encompassing core computation and data movements. Using this model, we systematically explore a wide parameter space, including intrinsic network characteristics (SNN time window size, spike rate, QNN sparsity, model size, weight bit-level) and hardware characteristics (memory system and network-on-chip). Our analysis identifies specific operational regimes where SNNs genuinely offer superior energy efficiency. For example, under typical neuromorphic hardware conditions, SNNs with moderate time windows ($T = 5$) require an average spike rate ($s_r$) below 5.7% to outperform equivalent QNNs These insights guide the design of truly energy-efficient neural network solutions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。