中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
In-Memory Computing Accelerator for Iterative Linear Algebra Solvers

文献类型:期刊论文

作者Liu, Rui1,2; Li, Zerun1,3; Zhang, Xiaoyu1,3; Chen, Xiaoming1,3; Han, Yinhe1,3; Tang, Minghua2
刊名IEEE COMPUTER ARCHITECTURE LETTERS
出版日期2025
卷号24期号:1页码:161-164
关键词System-on-chip Iterative methods Computer architecture Registers Delays Vectors In-memory computing Linear systems Data transfer Technological innovation Computing-in-memory iterative linear solver
ISSN号1556-6056
DOI10.1109/LCA.2025.3563365
英文摘要Iterative linear solvers are a crucial kernel in many numerical analysis problems. The performance and energy efficiency of iterative solvers based on traditional architectures are severely constrained by the memory wall bottleneck. Computing-in-memory (CIM) has the potential to enhance solving efficiency. Existing CIM architectures are mostly customized for specific algorithms and primarily focus on handling fixed-point operations, which makes them difficult to meet the demands of diverse and high-precision applications. In this work, we propose a CIM architecture that natively supports various iterative linear solvers based on floating-point operations. We develop a new instruction set for the accelerator, which can be flexibly combined to implement various iterative solvers. The evaluation results show that, compared with the GPU implementation, our accelerator achieves more than 10.1x speedup and 6.8x energy savings when executing different iterative solvers.
资助项目National Key R&D Program of China[2023YFF0719600] ; National Natural Science Foundation of China[62488101] ; National Natural Science Foundation of China[62495104] ; National Natural Science Foundation of China[62122076] ; National Natural Science Foundation of China[U23A20322] ; Youth Innovation Promotion Association CAS ; Hunan Provincial Natural Science Foundation[2023JJ50009] ; Hunan Provincial Natural Science Foundation[2023JJ30599] ; Foundation of Innovation Center of Radiation Application[KFZC2023020701] ; Postgraduate Scientific Research Innovation Project of Xiangtan University[XDCX2023Y184]
WOS研究方向Computer Science
语种英语
WOS记录号WOS:001498221200002
出版者IEEE COMPUTER SOC
源URL[http://119.78.100.204/handle/2XEOYT63/42321]  
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Chen, Xiaoming; Tang, Minghua
作者单位1.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
2.Xiangtan Univ, Sch Mat Sci & Engn, Xiangtan 411105, Peoples R China
3.Univ Chinese Acad Sci, Beijing 100190, Peoples R China
推荐引用方式
GB/T 7714
Liu, Rui,Li, Zerun,Zhang, Xiaoyu,et al. In-Memory Computing Accelerator for Iterative Linear Algebra Solvers[J]. IEEE COMPUTER ARCHITECTURE LETTERS,2025,24(1):161-164.
APA Liu, Rui,Li, Zerun,Zhang, Xiaoyu,Chen, Xiaoming,Han, Yinhe,&Tang, Minghua.(2025).In-Memory Computing Accelerator for Iterative Linear Algebra Solvers.IEEE COMPUTER ARCHITECTURE LETTERS,24(1),161-164.
MLA Liu, Rui,et al."In-Memory Computing Accelerator for Iterative Linear Algebra Solvers".IEEE COMPUTER ARCHITECTURE LETTERS 24.1(2025):161-164.

入库方式: OAI收割

来源:计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。