面向智算中心的新型以太网需求与关键技术

doi:10.11959/j.issn.1000-0801.2024171

Abstract

Abstract:

AI large model is leading the hot ICT(information and communications technology) industry in the next decade. Intelligent computing center network is a communication base to support the distributed training of AI large model, and it is one of the key factors to determine the efficiency of AI clusters. The data volume and the number of parameters of AI large model are expanding continuously, which brings the network of intelligent computing centers serious challenges, and also brings an opportunity for intergenerational innovation of key network technologies. In the process of AI large model training and inferencing, providing high performance and high security transmission of data are the two core requirements of AI business for intelligent computing network. Efficient load balancing, congestion control technologies and network security protocols are the key network technologies. To address the challenge brought by large-scale AI business, global scheduling ethernet (GSE) was proposed as a corresponding solution, and realistic test environment was built to compare the performance of GSE and RoCE. The test results show that GSE significantly improves JCT compared with RoCE network.

Key words: large model AI distributed training, GSE, load balancing, congestion control, network security protocol

CLC Number:

TP393

Xiaodong DUAN, Jieyu LI, Weiqiang CHENG, Han LI, Ruixue WANG, Haojie WANG. Challenges and key technologies of new Ethernet for intelligent computing center[J]. Telecommunications Science, 2024, 40(6): 146-159.

Figures/Tables 12

References 18

1	ZHANG Z, CHANG C K, LIN H B, et al. "Is network the bottleneck of distributed training?"[EB]. arXiv preprint arXiv, 2020: 2006.10103.
2	LUO L, WEST P, KRISHNAMURTHY A, CEZE L, et al. PLink: discovering and exploiting datacenter network locality for efficient cloud-based distributed training[C]//Proceedings of 2020 MLSys. 2020.
3	ROTHENBERGER B, TARANOV K, PERRIG A, et al. {ReDMArk}: Bypassing {RDMA} security mechanisms[C]//30th USENIX Security Symposium (USENIX Security 21). 2021: 4277-4292.
4	HOEFLER T, ROWETH D, UNDERWOOD K, et al. Datacenter Ethernet and RDMA: issues at hyperscale[EB]. arXiv preprint arXiv, 2023: 2302.03337.
5	HOPPS C. Analysis of an equal-cost multi-path algorithm[R]. Technical Report, 2000.
6	QURESHI M A, CHENG Y C, YIN Q W,et al. PLB: congestion signals are simple and effective for network load balancing[C]// Proceedings of the ACM SIGCOMM 2022 Conference. New York: ACM Press, 2022: 207-218.
7	SONG C H, KHOOI X Z, JOSHI R, et al. Network load balancing with in-network reordering support for RDMA[C]//Proceedings of the ACM SIGCOMM 2023 Conference. 2023: 816-831.
8	DIXIT A, PRAKASH A, HU Y C, et al. On the impact of packet spraying in data center networks[C]//Proceedings of IEEE INFOCOM 2013. Piscataway: IEEE Press, 2013: 2130-2138.
9	SCHARF M, KIESEL S. NXG03-5: Head-of-line blocking in TCP and SCTP: analysis and measurements[C]//Proceedings of the 49th IEEE Global Telecommunications Conference(GLOBECOM 2006). Piscataway: IEEE Press, 2006: 1-5.
10	XUE J C, CHAUDHRY M U, VAMANAN B, et al. Dart: divide and specialize for fast response to congestion in RDMA-based datacenter networks[J]. IEEE/ACM Transactions on Networking, 2020, 28(1): 322-335.
11	ZHU Y B, ERAN H, FIRESTONE D, et al. Congestion control for large-scale RDMA deployments[J]. ACM SIGCOMM Computer Communication Review, 2015, 45(4): 523-536..
12	HU S H, ZHU Y B, CHENG P, et al. Deadlocks in datacenter networks: why do they form, and how to avoid them[C]//Proceedings of the 15th ACM Workshop on Hot Topics in Networks. New York: ACM Press, 2016: 92-98.
13	MITTAL R, THE L V, DUKKIPATI N, et al. TIMELY: RTT-based congestion control for the datacenter[C]//Proceedings of SIGCOMM 2015. 2015: 537-550.
14	LI Y, MIAO R, LIU H H, et al. HPCC: high precision congestion control[C]//Proceedings of the ACM Special Interest Group on Data Communication. NewYork: ACM Press, 2019: 44-58.
15	PINKERTON J, DELEGANES E. RFC 5042:direct data placement protocol (DDP)/remote direct memory access protocol (RDMAP) security[R]. IEFT, 2007.
16	Google. Google white paper: PSP architecture specification[R]. 2022.
17	IEEE 802.1AE-2018: media access control (MAC) security [S]. 2018.
18	HOPPS C. RFC 9347:aggregation and fragmentation mode for encapsulating security payload (ESP) and its use for IP traffic flow security (IP-TFS)[R]. IEFT, 2023.

Metrics

Recommended 0

No Suggested Reading articles found!

[1]	Wenlei GUO. Optimization model of random load balancing of heterogeneous nodes in complex buildings [J]. Telecommunications Science, 2023, 39(7): 139-148.
[2]	Yu DING, Chenkai LI, Huimei HAN, Weidang LU, Yuanhong REN, Yuan GAO, Jiang CAO. Multi-agent heterogeneous network selection method based on 5G UAV communication [J]. Telecommunications Science, 2022, 38(8): 28-36.
[3]	Zhaohui WANG, Jianliang SHEN, Xia ZHANG, Ting CHEN. FALB: a FC protocol link aggregation algorithm [J]. Telecommunications Science, 2021, 37(9): 105-111.
[4]	Sining LUO,Hualong WANG,Hongyu LI,Wei PENG. Improved ant colony algorithm based cloud computing user task scheduling algorithm [J]. Telecommunications Science, 2020, 36(2): 95-100.
[5]	Guangxue YUE,Yasheng DAI,Xiaohui YANG,Youkang ZHU,Zhenxu YOU,Jiansheng LIU. Multi-constrained trusted cooperative task migration strategy for edge computing [J]. Telecommunications Science, 2019, 35(11): 36-50.
[6]	Haipeng LIU. A static MEH handover scheme based on load balancing [J]. Telecommunications Science, 2018, 34(7): 118-127.
[7]	Wenbo WANG,Qingwei YE,Yu ZHOU,Zhihua LU. Dynamic load balancing algorithm based on queuing theory comprehensive index evaluation [J]. Telecommunications Science, 2018, 34(7): 86-91.
[8]	Qiao SUN,Buqiao DENG,Zhiqiang WANG,Xubin PEI. A scalable load balancing strategy based on distributed server cluster [J]. Telecommunications Science, 2017, 33(9): 190-196.
[9]	Xiaoan BAO,Xue WEI,Lei CHEN,Guoheng HU,Na ZHANG. Load balancing method of service cluster based on mean-variance [J]. Telecommunications Science, 2017, 33(1): 1-8.
[10]	Nan Bao,Weiwei Xia,Xu Bao. Network Selection Algorithm Based on Weighted Bipartite Graph in Heterogeneous Network Environment [J]. Telecommunications Science, 2015, 31(9): 44-50.
[11]	Kai Liu,Jianye Zhao,Zhimin Liu. Power Efficient Load Balancing Algorithm in LTE-A Heterogenous Network Based on Assignment Model [J]. Telecommunications Science, 2015, 31(2): 22-28.
[12]	Zhendong Zhao,Weijiao Niu,Dongsheng Han,Shasha Ding. Cell Selection Algorithm for Edge Users in Heterogeneous Wireless Network [J]. Telecommunications Science, 2015, 31(12): 52-57.
[13]	Junrong Yan,Yulong Lu,Peng Pan. An Error Tolerated Congestion Control Algorithm Based on Packet Loss Rate Measurement in Deep Space Communication Network [J]. Telecommunications Science, 2015, 31(11): 43-47.
[14]	Yue Liu,Xiaolei Guo,Xun Zhang. Network Optimization Algorithm Based on Power Control and Channel Assignment [J]. Telecommunications Science, 2015, 31(10): 93-98.
[15]	Kai Lei,Jie Yuan. Content Distribution Mecbanisms of Named Data Networking [J]. Telecommunications Science, 2014, 30(9): 27-36.

Challenges and key technologies of new Ethernet for intelligent computing center

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 12

References 18

Related Articles 15

Metrics

Recommended 0