arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.08044cs.AR
收藏

面向基于3D-DRAM的LLM加速器的全栈性能评估基础设施

A Full-Stack Performance Evaluation Infrastructure for 3D-DRAM-based LLM Accelerators

Peking University(北京大学);Xi’an UniIC Semiconductors(西安联芯半导体);Houmo AI(厚墨智能);HKUST(香港科技大学)

Cong Li, Chenhao Xue, Yi Ren, Xiping Dong, Yu Cheng, Yinbo Hu, Fujun Bai, Yixin Guo, Xiping Jiang, Qiang Wu, Zhi Yang, Zhe Cheng, Yuan Xie, Guangyu Sun

AI总结:

本文提出ATLAS框架,通过统一抽象支持任意LLM推理场景,验证其在真实硅片上的高精度模拟性能,为3D-DRAM内存系统和加速器微架构设计提供指导。

摘要:

大型语言模型(LLM)在解码过程中表现出内存密集型行为,成为LLM推理的关键瓶颈。为加速解码执行,混合键合的3D-DRAM已被应用于LLM加速器。尽管该新兴技术在现有硬件上提供了显著性能提升,但当前的3D-DRAM加速器(3D-加速器)依赖于闭源评估工具,限制了公开可用的性能分析方法。此外,现有设计高度定制化,缺乏适用于多样化用例的通用且可重用的全栈建模。为弥合这一根本性差距,本文提出了ATLAS,首个经过硅片验证的基于3D-DRAM的LLM加速器模拟框架。基于商用部署的多层3D-DRAM技术,ATLAS引入统一抽象,支持任意LLM推理场景。验证显示,ATLAS的模拟误差小于等于8.57%,与测量性能的相关性达97.26-99.96%。通过ATLAS的设计空间探索,展示了其指导架构设计和提取3D-DRAM内存系统及3D-加速器微架构关键结论的能力。ATLAS将在发表后开源,促进进一步的3D-加速器研究。

英文摘要:

Large language models (LLMs) exhibit memory-intensive behavior during decoding, making it a key bottleneck in LLM inference. To accelerate decoding execution, hybrid-bonding-based 3D-DRAM has been adopted in LLM accelerators. While this emerging technology provides strong performance gains over existing hardware, current 3D-DRAM accelerators (3D-Accelerators) rely on closed-source evaluation tools, limiting access to publicly available performance analysis methods. Moreover, existing designs are highly customized for specific scenarios, lacking a general and reusable full-stack modeling for 3D-Accelerators across diverse usecases. To bridge this fundamental gap, we present ATLAS, the first silicon-proven Architectural Three-dimesional-DRAM-based LLM Accelerator Simulation framework. Built on commercially deployed multi-layer 3D-DRAM technology, ATLAS introduces unified abstractions for both 3D-Accelerator system architecture and programming primitives to support arbitrary LLM inference scenarios. Validation against real silicon shows that ATLAS achieves $\le$8.57% simulation error and 97.26-99.96\% correlation with measured performance. Through design space exploration with ATLAS, we demonstrate its ability to guide architecture design and distill key takeaways for both 3D-DRAM memory system and 3D-Accelerator microarchitecture across scenarios. ATLAS will be open-sourced upon publication, enabling further research on 3D-Accelerators.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑