Unleashing Low-Bit Inference on Ascend NPUs: A Comprehensive Evaluation of HiFloat Formats

Published in ACL, 2026

Recommended citation: Pengxiang Zhao, Hui-Ling Zhen, Xing Li, Han Bao, Weizhe Lin, Zhiyuan Yang, Yu Zi Wei, Xin Wang, Mingxuan Yuan, Xianzhi Yu, Zhenhua Dong. (2026). Unleashing Low-Bit Inference on Ascend NPUs: A Comprehensive Evaluation of HiFloat Formats. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track), 844–864. https://doi.org/10.18653/v1/2026.acl-industry.57 https://aclanthology.org/2026.acl-industry.57/