Metraipolis › archive › dc_c60eaef50418
HBQ: Hierarchical Scaling Block Quantization with Hardware-Efficiency-Aware Design for Accurate LLM Inference — arXiv:2609.00450
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00450
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:33:23.030Z
- SHA-256 of the archived text
d2898bcade16a83af2c753e50177ee4340fb18ff35aa9e7a0c22e1eafa497f21- Stable URL of this page
https://metraipolis.com/documento/dc_c60eaef50418
How to cite
HBQ: Hierarchical Scaling Block Quantization with Hardware-Efficiency-Aware Design for Accurate LLM Inference — arXiv:2609.00450 — arXiv (en), https://arxiv.org/abs/2609.00450 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c60eaef50418 (Metraipolis archive dc_c60eaef50418, sha256 d2898bcade16a83af2c753e50177ee4340fb18ff35aa9e7a0c22e1eafa497f21)
BibTeX
@misc{dcc60eaef50418,
title = {HBQ: Hierarchical Scaling Block Quantization with Hardware-Efficiency-Aware Design for Accurate LLM Inference — arXiv:2609.00450},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00450},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c60eaef50418; sha256 d2898bcade16a83af2c753e50177ee4340fb18ff35aa9e7a0c22e1eafa497f21},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00450. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Block Quantization (BQ) enables efficient LLM inference by quantizing both weights and activations, but its design space remains underexplored. Through hardware-accuracy design space exploration, we identify block size as a key trade-off: larger blocks improve hardware efficiency by amortizing dequantization and accumulation costs, but degrade accuracy. Motivated by this insight, we propose Hierarchical Block Quantization (HBQ), which combines large blocks with low-overhead significand (SIG) scaling for second-level quantization. SIG scaling effectively compensates for large-block quantization errors while accounting for distinct weight and activation distributions. HBQ-A achieves W4A16-level accuracy with W4A5 and lower area than NVFP4, while HBQ-E further reduces hardware cost by 17% while outperforming existing BQ methods in accuracy. We implement HBQ for weights, activations, and KV cache in a 28nm ASIC accelerator and introduce partial-sum BQ to reduce EMA energy. At comparable accuracy, HBQ achieves 2.3x/4.6x higher area/energy efficiency than state-of-the-art weight-only quantization and 1.6-3.3x lower system energy with 1.5-3x speedup over prior BQ methods. Our implementation is publicly available at: this https URL . Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR). Pagina del lavoro: https://arxiv.org/abs/2609.00450 — 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 2026-10-17T06:15:12.783Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).