SLAM-CIM: A Visual SLAM Backend Processor With Dynamic-Range-Driven-Skipping Linear-Solving FP-CIM Macros

doi:10.1109/JSSC.2024.3402808

	SLAM-CIM: A Visual SLAM Backend Processor With Dynamic-Range-Driven-Skipping Linear-Solving FP-CIM Macros
	Li, Mengjie 1; Zhu, Haozhe 1; He, Siqi 1; Zhang, Hongyi 1; Liao, Jie 1; Zhai, Danfeng 1; Chen, Chixiao 1; Liu, Qi 1; Zeng, Xiaoyang 1; Sun, Ninghui 2; Liu, Ming 1
	2024-05-30
发表期刊	IEEE JOURNAL OF SOLID-STATE CIRCUITS
ISSN	0018-9200
页码	13
摘要	Simultaneous localization and mapping (SLAM), a pivotal technology in robotics, autonomous vehicles, and surveillance, has gained prominence with the emergence of edge intelligence. Developing energy-efficient, low-latency SLAM systems is essential due to resource constraints and real-time demands. Compute-in-memory (CIM) architectures have been proven to be efficient for matrix multiplications. However, applications for SLAM raise new challenges in memory access and computation aspects: the linear system solving (LS) requires row transformation and causes frequent CIM updates, while the backend optimization causes redundant memory access; back-end optimization dominates SLAM's computation and requires high precision and high dynamic range. Thus, we propose SLAM-CIM, a visual SLAM backend processor for edge robotics. A dynamic-range-driven-skipping CIM macro is designed to realize energy-efficient floating point (FP)-multiply-and-accumulate (MAC) operations. A preconditional-conjugate-gradient-based in-memory linear solver (PILARS) is designed to achieve LS without additional row transformations. This reduces memory access by 2.08 $\times$ and linear-system-solving latency by 3.84 x. SLAM-CIM further minimizes CIM weight updates through incremental bundle adjustment (BA), increasing average CIM utilization by 2.8 x. A silicon prototype is fabricated using 28-nm CMOS technology. The measurements show that SLAM-CIM achieves accurate and efficient SLAM operations with an average energy efficiency of 31.53 TFLOPS/W.
关键词	Simultaneous localization and mapping Sorting In-memory computing Visualization Energy efficiency Optimization Common Information Model (computing) Compute in memory (CIM) floating point (FP) linear system solver simultaneous localization and mapping (SLAM)
DOI	10.1109/JSSC.2024.3402808
收录类别	SCI
语种	英语
资助项目	Major Project of the Science and Technology Innovation 2030
WOS研究方向	Engineering
WOS类目	Engineering, Electrical & Electronic
WOS记录号	WOS:001236627600001
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40051
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Zhu, Haozhe; Chen, Chixiao
作者单位	1.Fudan Univ, State Key Lab Integrated Chips & Syst, Shanghai 200433, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Li, Mengjie,Zhu, Haozhe,He, Siqi,et al. SLAM-CIM: A Visual SLAM Backend Processor With Dynamic-Range-Driven-Skipping Linear-Solving FP-CIM Macros[J]. IEEE JOURNAL OF SOLID-STATE CIRCUITS,2024:13.
APA	Li, Mengjie.,Zhu, Haozhe.,He, Siqi.,Zhang, Hongyi.,Liao, Jie.,...&Liu, Ming.(2024).SLAM-CIM: A Visual SLAM Backend Processor With Dynamic-Range-Driven-Skipping Linear-Solving FP-CIM Macros.IEEE JOURNAL OF SOLID-STATE CIRCUITS,13.
MLA	Li, Mengjie,et al."SLAM-CIM: A Visual SLAM Backend Processor With Dynamic-Range-Driven-Skipping Linear-Solving FP-CIM Macros".IEEE JOURNAL OF SOLID-STATE CIRCUITS (2024):13.