DeInfer: Efficient Parallel Inferencing for Decomposed Large Language Models Conference: 63rd ACM/IEEE Design Automation Conference; July 26–29, 2026; Long Beach, CA, USA 63rd ACM/IEEE Design Automation Conference (DAC ’26), July 26–29, 2026, Long Beach, CA, USA DOI: 10.1145/3770743.3804360 ISBN: 979-8-4007-2254-7/2026/07 CCS: Computing methodologies Neural networks CCS: Computing methodologies Parallel algorithms CCS: Computing methodologies Natural language processing

You-Liang Huang, Xinhao Huang, Chengxi Liao, Zeyi Wen · 2026

Read DeInfer: Efficient Parallel Inferencing for Decomposed Large Language Models Conference: 63rd ACM/IEEE Design Automation Conference; July 26–29, 2026; Long Beach, CA, USA 63rd ACM/IEEE Design Automation Conference (DAC ’26), July 26–29, 2026, Long Beach, CA, USA DOI: 10.1145/3770743.3804360 ISBN: 979-8-4007-2254-7/2026/07 CCS: Computing methodologies Neural networks CCS: Computing methodologies Parallel algorithms CCS: Computing methodologies Natural language processing by You-Liang Huang, Xinhao Huang, Chengxi Liao, Zeyi Wen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-17709.txt.

Description

Huang and three colleagues introduce DeInfer, a fast system for side-by-side runs of decomposed large language models. Old decomposition work focuses on downstream task results but ignores poor side-by-side runs when scaling up model size. DeInfer combines several tweaks to lift speed while working with leading techniques. Broad experiments show its edge, suggesting it can greatly ease side-by-side runs of decomposed LLMs.

Catalog record

CREATORYou-Liang Huang, Xinhao Huang, Chengxi Liao, Zeyi Wen
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE31821
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.17709

Browse the full library · About the Tower of Babel