Metraipolis › archive › dc_b3188b0bccfe

A Feature-Major Codebook for Memory-Efficient Sparse-Binary Self-Organizing Maps: Scaling a MEDLINE Atlas to 1.05 Million Neurons on a Single Consumer GPU — arXiv:2608.24067

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.24067
License (as archived)
CC0-1.0
Archived
2026-10-18T06:22:03.308Z
SHA-256 of the archived text
78f3bdda53db826b6efd84361975bae5da368dcceb4944bb7266c0ca67578846
Stable URL of this page
https://metraipolis.com/documento/dc_b3188b0bccfe

How to cite

A Feature-Major Codebook for Memory-Efficient Sparse-Binary Self-Organizing Maps: Scaling a MEDLINE Atlas to 1.05 Million Neurons on a Single Consumer GPU — arXiv:2608.24067 — arXiv (en), https://arxiv.org/abs/2608.24067 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b3188b0bccfe (Metraipolis archive dc_b3188b0bccfe, sha256 78f3bdda53db826b6efd84361975bae5da368dcceb4944bb7266c0ca67578846)
BibTeX
@misc{dcb3188b0bccfe,
  title = {A Feature-Major Codebook for Memory-Efficient Sparse-Binary Self-Organizing Maps: Scaling a MEDLINE Atlas to 1.05 Million Neurons on a Single Consumer GPU — arXiv:2608.24067},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.24067},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b3188b0bccfe; sha256 78f3bdda53db826b6efd84361975bae5da368dcceb4944bb7266c0ca67578846},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.24067. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Building a self-organising map at MEDLINE scale has been impractical: the best-matching-unit (BMU) search that dominates training is bound by the bandwidth needed to read the codebook every epoch. I show that this bottleneck is largely an artefact of codebook layout. Storing it feature-major with each feature's weights contiguous, W[v.M+i], recasts the search as a tiled sparse-dense product in which every loaded weight column is reused across a tile of samples. Varying only the layout, with implementation, precision and update rule held fixed, accelerates the BMU search by 4.5-8.5x, and because an exact-argmin BMU is invariant to codebook layout, this costs nothing: held-out quantisation error agrees with a cuSPARSE baseline to within 0.5% at every map size. The advantage is a crossover: cuSPARSE$.$SOM is faster at small maps, SparseBin$.$SOM is 1.5x faster at 128x128 and 2.6x at 256x256, and at 512x512 it is the only one that runs on 24 GB without re-engineering its memory path. Paired with a radius-independent box-blur update and a convergence-based stopping rule, it trains a converged map over 29.9 million MEDLINE articles in about 72 s at 64x64 on one 24 GB GPU, and fits 262,144 neurons (512x512) where every alternative I tested exceeds memory; on a 141 GB H200 it reaches 1,048,576 neurons (1024x1024), to my knowledge the largest self-organising map yet reported. Held-out error follows a smooth power law with no elbow across three decades of map size. At matched work, in the configuration benchmarked here, the design is ~82x faster than MedSOM and, at 128x128, 621x faster than the best multicore-CPU library. A post-submission addendum, tuning both implementations symmetrically, accelerates the search a further 5.6-10.1x, brings that 64x64 run to about 13 s, removes the crossover, raises those margins to ~385x and ~3,000x, and narrows two mechanism claims.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-08-26 in cs.DC).
Categorie arXiv: Machine Learning (cs.LG) ; Distributed, Parallel, and Cluster Computing (cs.DC).
Pagina del lavoro: https://arxiv.org/abs/2608.24067 — 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 2026-10-18T06:17:57.837Z dalla pagina https://arxiv.org/catchup/cs.DC/2026-08-26?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).