AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

HardF-SNN: Hardware-Friendly Quantization for Spiking Neural Networks with Efficient Integer-Arithmetic-Only Inference

Hanwen Liu, Kexin Shi, Jieyuan Zhang, Yimeng Shan, Jibin Wu, Wenyu Chen, Malu Zhang

AAAI 2026 · Architectures and efficiency › Spiking networks · approach: Training deep spiking networks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add proportional shared-scale quantization (PSSQ) to the baseline (ablation)HardF-SNN baseline (ResNet19, 4w-4u) · CIFAR-100accuracy (%): 73.15 -> 74.93Improved
add proportional shared-scale quantization (PSSQ) to the baseline (ablation)HardF-SNN baseline (ResNet19, 8w-8u) · CIFAR-100accuracy (%): 74.82 -> 75.62Improved
add integer-only BN (I-BN) to the baseline (ablation)HardF-SNN baseline (ResNet19, 8w-8u) · CIFAR-100accuracy (%): 74.82 -> 75.04Improved
add integer-only BN (I-BN) to the baseline (ablation)HardF-SNN baseline (ResNet19, 4w-4u) · CIFAR-100accuracy (%): 73.15 -> 74.61Improved
add integer-only BN so 2-bit training converges (ablation)HardF-SNN baseline (ResNet19, 2w-2u) · CIFAR-100trainability / accuracy (%): untrainable -> 73.93Improved
add PSSQ on top of I-BN (ablation)HardF-SNN (ResNet19, 2w-2u) · CIFAR-100accuracy (%): 73.93 -> 74.22Improved
add both PSSQ and I-BN (ablation)HardF-SNN baseline (ResNet19, 4w-4u) · CIFAR-100accuracy (%): 73.15 -> 75.35Improved
use power-of-two learnable eta instead of rounded max-abs eta (ablation)PSSQ (ResNet19) · CIFAR-100accuracy (%)No clear effect
reduce bit-width from 8w-8u to 2w-2u (ablation)HardF-SNN (ResNet18) · ImageNet-1kaccuracy (%): 59.52 -> 57.71Hurt
reduce bit-width from 8w-8u to 2w-2u (ablation)HardF-SNN (ResNet19) · CIFAR-10accuracy (%): 95.47 -> 94.49Hurt
lower FPGA precision from 8-bit to 2-bit (ablation)HardF-SNN accelerator (VGG11) · CIFAR-100 FPGAaccuracy (%) / throughput (GOPS): 67.26 -> 65.49 accuracy; 297.44 -> 993.16 GOPSMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy