Reproduce “STISA: A 0.16 GOPS/W/PE Single-Shot Inference FPGA-based SNN Accelerator with Algorithm and Hardware Co-Design” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Kainan Wang, Chengyi Yang, Lei Liu, Chengting Yu, Yee Sin Ang, Bo Wang, Aili Wang

STISA co-designs block-level temporal compression with an FPGA accelerator for spiking neural network image classification. Its Temporal Splitter Compression method assigns different timestep budgets across blocks, while Temporal-Prioritized Output-Stationary dataflow and Block-Mux Streaming support execution. On CIFAR-10 ResNet19, it reduces operations 51.71% with a 0.29-point accuracy drop; ImageNet ResNet34 hardware gains apply to the reported configurations, not universally.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.