mlsanityData to 7 Oct 2026

Reconsidering the Energy Efficiency of Spiking Neural Networks Inference from Analytical Perspectives

arXiv 2409.08290 ↗First posted 29 Aug 2024cs.NE cs.AI cs.LGcited by 12

Not on the map. Why?

Abstract (arXiv)

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 log2(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 (sr) below 5.7% to outperform equivalent QNNs These insights guide the design of truly energy-efficient neural network solutions.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 23 of its 83 references that are in the map. First posted on arXiv 29 Aug 2024.

Builds on

23 of 83 references.

References · 23

  1. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  2. 2021RepVGG: Making VGG-style ConvNets Great Againcited by 520
  3. 2017Efficient Processing of Deep Neural Networks: A Tutorial and Surveycited by 427
  4. 2021Training Spiking Neural Networks Using Lessons From Deep Learningcited by 251
  5. 2020RMP-SNN: Residual Membrane Potential Neuron for Enabling Deeper High-Accuracy and Low-Latency Spiking Neural Networkcited by 141
  6. 2023SpikingBERT: Distilling BERT to Train Spiking Language Models Using Implicit Differentiationcited by 47
  7. 2017STDP Based Pruning of Connections and Weight Quantization in Spiking Neural Networks for Energy Efficient Recognitioncited by 41
  8. 2022The BrainScaleS-2 accelerated neuromorphic system with hybrid plasticitycited by 41
  9. 2024SpikeLM: Towards General Spike-Driven Language Modeling via Elastic Bi-Spiking Mechanismscited by 31
  10. 2018Neural Cache: Bit-Serial In-Cache Acceleration of Deep Neural Networkscited by 29
  11. 2024SpikeLLM: Scaling up Spiking Neural Network to Large Language Models via Saliency-based Spikingcited by 29
  12. 2021Dual-side Sparse Tensor Corecited by 27
11 more
  1. 2022SATA: Sparsity-Aware Training Accelerator for Spiking Neural Networkscited by 18
  2. 2024Event-Driven Learning for Spiking Neural Networkscited by 15
  3. 2023To Spike or Not To Spike: A Digital Hardware Perspective on Deep Learning Accelerationcited by 10
  4. 2022SpikeSim: An end-to-end Compute-in-Memory Hardware Evaluation Tool for Benchmarking Spiking Neural Networkscited by 8
  5. 2024LoAS: Fully Temporal-Parallel Dataflow for Dual-Sparse Spiking Neural Networkscited by 7
  6. 2023Are SNNs Truly Energy-efficient? - A Hardware Perspectivecited by 6
  7. 2022Low Latency Conversion of Artificial Neural Network Models to Rate-encoded Spiking Neural Networkscited by 5
  8. 2024Fast and Controllable Post-training Sparsity: Learning Optimal Sparsity Allocation with Global Constraint in Minutescited by 3
  9. 2020SNEAP: A Fast and Efficient Toolchain for Mapping Large-Scale Spiking Neural Network onto NoC-based Neuromorphic Platformcited by 2
  10. 2024Allspark: Workload Orchestration for Visual Transformers on Processing In-Memory Systemscited by 2
  11. 2024SparrowSNN: A Hardware/software Co-design for Energy Efficient ECG Classificationcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 5.

12 papers.

  1. 2025Otters: An Energy-Efficient SpikingTransformer via Optical Time-to-First-Spike Encodingcited by 5
  2. 2025Energy-Efficient and Dequantization-Free Q-LLMs: A Spiking Neural Network Approach to Salient Value Mitigationcited by 2
  3. 2025Spiking Neural Networks: The Future of Brain-Inspired Computingcited by 2
  4. 2025LightSNN: Lightweight Architecture Search for Sparse and Accurate Spiking Neural Networkscited by 1
  5. 2025Hardware-aware vs. Hardware-agnostic Energy Estimation for SNN in Space Applicationscited by 1
  6. 2026Kirin: Improving ANN efficiency with SNN Hybridizationcited by 1
  7. 2025Spiking Neural Networks for SAR Interferometric Phase Unwrapping: A Theoretical Framework for Energy-Efficient Processingcited by 0
  8. 2025Random Feature Spiking Neural Networkscited by 0
  9. 2026SpikySpace: A Spiking State Space Model for Energy-Efficient Time Series Forecastingcited by 0
  10. 2026UltraLIF: Fully Differentiable Spiking Neural Networks via Ultradiscretization and Max-Plus Algebracited by 0
  11. 2026Otters++: A Time-to-first-spike Based Energy Efficient Optical Spiking Transformercited by 0
  12. 2026Latency and accuracy tradeoffs in Spiking Neural Networkscited by 0