AMALI: An Analytical Model for Accurately Modeling LLM Inference on Modern GPUs

Shiheng CaoJunmin WuJunshi ChenHong AnZhibin Yu

体系结构与近数据计算方向论文：AMALI: An Analytical Model for Accurately Modeling LLM Inference on Mode

↗在线页面Official page ◇DOI

内容导读Reading note

分类与摘要Category & abstract

从处理器、指令、内存或近数据计算角度优化科学工作负载。

Metadata

完整论文信息Full record

title: AMALI: An Analytical Model for Accurately Modeling LLM Inference on Modern GPUs
title_zh: 待补充
abstract: 待补充
abstract_zh: 待补充
keywords: GPU, AI, Architecture, 学位认定 A, CCF A
year: 2025
published_date: 待补充
online_date: 待补充
paper_type: Conference
publication_status: Published
volume: 待补充
issue: 待补充
pages: 待补充
article_number: 待补充
publisher: 待补充
doi: 10.1145/3695053.3731064
research_area: 体系结构与近数据计算
tags: GPU, AI, Architecture, 学位认定 A, CCF A
category: 体系结构与近数据计算
summary: 从处理器、指令、内存或近数据计算角度优化科学工作负载。
authors: Shiheng Cao, Junmin Wu, Junshi Chen, Hong An, Zhibin Yu
corresponding_authors: 待补充
affiliations: 待补充
funding: 待补充

Research area

体系结构与近数据计算

Citation

Shiheng Cao, Junmin Wu, Junshi Chen, Hong An, and Zhibin Yu. AMALI: An Analytical Model for Accurately Modeling LLM Inference on Modern GPUs. In Proceedings of the 52nd Annual International Symposium on Computer Architecture (ISCA '25). Tokyo, Japan, June 21– 25, 2025. 1495-1508 (CCF A)

BibTeX

@article{acsa2025_11,
  title = {AMALI: An Analytical Model for Accurately Modeling LLM Inference on Modern GPUs},
  year = {2025},
  doi = {10.1145/3695053.3731064}
}