Metraipolis › archive › dc_c6bbc604d44f

PSA: A Pruning-Friendly Sparse-Aware Accelerator for Efficient Sparse Convolution in SNNs — arXiv:2610.21313

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.21313
License (as archived)
CC0-1.0
Archived
2026-10-16T10:04:53.993Z
SHA-256 of the archived text
f6c726f4836a9e8af035b52c725dfe3a927ef3d4641a6ecadbab32709f01511e
Stable URL of this page
https://metraipolis.com/documento/dc_c6bbc604d44f

How to cite

PSA: A Pruning-Friendly Sparse-Aware Accelerator for Efficient Sparse Convolution in SNNs — arXiv:2610.21313 — arXiv (en), https://arxiv.org/abs/2610.21313 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c6bbc604d44f (Metraipolis archive dc_c6bbc604d44f, sha256 f6c726f4836a9e8af035b52c725dfe3a927ef3d4641a6ecadbab32709f01511e)
BibTeX
@misc{dcc6bbc604d44f,
  title = {PSA: A Pruning-Friendly Sparse-Aware Accelerator for Efficient Sparse Convolution in SNNs — arXiv:2610.21313},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.21313},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c6bbc604d44f; sha256 f6c726f4836a9e8af035b52c725dfe3a927ef3d4641a6ecadbab32709f01511e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.21313. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Spiking neural networks (SNNs) are promising for energy-efficient artificial intelligence due to their event-driven sparse computation, and neuron pruning can further reduce redundant operations. However, in practical hardware accelerators, spike sparsity and pruning-induced sparsity produce highly irregular sparse workloads that are difficult to execute efficiently. Existing approaches do not jointly support sparse-input execution, low-overhead sparse convolution lowering and pruning-friendly execution. This paper presents PSA, a Pruning-friendly Sparse-aware Accelerator for efficient sparse convolution in SNNs. PSA features a Sparse-Aware Merge-Accumulation (SAMA) dataflow that aligns sparse spike indices and converts irregular activations into continuous computation streams for efficient sparse convolution execution. It further introduces a Sparse Implicit Group-wise im2col (SIG-im2col) convolution lowering mechanism that converts sparse convolution into SpMM on-chip with low mapping cost and memory overhead. In addition, PSA integrates a lightweight pruning-friendly gating mechanism to directly support pruned SNNs. Results on ResNet-18/CIFAR-10 show that PSA achieves a 4.45x improvement in energy efficiency and an 18.52x reduction in energy-delay product (EDP) over the dense baseline.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.AR, nuovi lavori del giorno).
Categorie arXiv: Hardware Architecture (cs.AR).
Pagina del lavoro: https://arxiv.org/abs/2610.21313 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T08:45:17.649Z dalla pagina https://arxiv.org/list/cs.AR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).