DPU-Direct: Unleashing Remote Accelerators via Enhanced RDMA for Disaggregated Datacenters

doi:10.1109/TC.2024.3404089

	DPU-Direct: Unleashing Remote Accelerators via Enhanced RDMA for Disaggregated Datacenters
	Liao, Yunkun 1,2,3; Wu, Jingya 1; Lu, Wenyan 1,4; Li, Xiaowei 3; Yan, Guihai 1,4
	2024-08-01
发表期刊	IEEE TRANSACTIONS ON COMPUTERS
ISSN	0018-9340
卷号	73 期号:8 页码:2081-2095
摘要	This paper presents DPU-Direct, an accelerator disaggregation system that connects accelerator nodes (ANs) and CPU nodes (CNs) over a standard Remote Direct Memory Access (RDMA) network. DPU-Direct eliminates the latency introduced by the CPU-based network stack, and PCIe interconnects between network I/O and the accelerator. The DPU-Direct system architecture includes a DPU Wrapper hardware architecture, an RDMA-based Accelerator Access Pattern (RAAP), and a CN-side programming model. The DPU Wrapper connects accelerators directly with the RDMA engine, turning ANs into disaggregation-native devices. The RAAP provides the CN with low-latency and high throughput accelerator semantics based on standard RDMA operations. Our FPGA prototype demonstrates DPU-Direct's efficacy with two proof-of-concept applications: AES encryption and key-value cache, which are computationally intensive and latency-sensitive. DPU-Direct yields a 400x speedup in AES encryption over the CPU baseline and matches the performance of the locally integrated AES accelerator. For key-value cache, DPU-Direct reduces the average end-to-end latency by 1.66x for GETs and 1.30x for SETs over the CPU-RDMA-Polling baseline, reducing latency jitter by over 10x for both operations.
关键词	Central Processing Unit Engines Jitter Computers Pipelines Programming Encryption Disaggregated datacenter SmartNIC RDMA hardware accelerator
DOI	10.1109/TC.2024.3404089
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China (NSFC)[62002340] ; National Natural Science Foundation of China (NSFC)[62090020] ; National Natural Science Foundation of China (NSFC)[61872336] ; Youth Innovation Promotion Association CAS[Y201923] ; Strategic Priority Research Program of the Chinese Academy of Sciences[XDB44030100] ; Internship program of YUSUR Technology Co., Ltd.
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Hardware & Architecture ; Engineering, Electrical & Electronic
WOS记录号	WOS:001270596400013
出版者	IEEE COMPUTER SOC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/39839
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Wu, Jingya; Yan, Guihai
作者单位	1.Chinese Acad Sci, Inst Comp Technol, SKLP, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100190, Peoples R China 3.Zhongguancun Lab, Beijing 100190, Peoples R China 4.YUSUR Tech Co Ltd, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Liao, Yunkun,Wu, Jingya,Lu, Wenyan,et al. DPU-Direct: Unleashing Remote Accelerators via Enhanced RDMA for Disaggregated Datacenters[J]. IEEE TRANSACTIONS ON COMPUTERS,2024,73(8):2081-2095.
APA	Liao, Yunkun,Wu, Jingya,Lu, Wenyan,Li, Xiaowei,&Yan, Guihai.(2024).DPU-Direct: Unleashing Remote Accelerators via Enhanced RDMA for Disaggregated Datacenters.IEEE TRANSACTIONS ON COMPUTERS,73(8),2081-2095.
MLA	Liao, Yunkun,et al."DPU-Direct: Unleashing Remote Accelerators via Enhanced RDMA for Disaggregated Datacenters".IEEE TRANSACTIONS ON COMPUTERS 73.8(2024):2081-2095.