Five forces of computational power: a comprehensive method to measure computational power

doi:10.12267/j.issn.2096-5931.2022.03.003

Abstract

Abstract:

In this paper, the “Five Forces of Computational Power” is proposed to comprehensively evaluate the computational power of data centers. The model solves the problem that the measurement units of different types of indicators are not unified in the process of computational power evaluation, and combines the bidirectional projection method and TOPSIS method to evaluate the comprehensive situation of computational power more scientifically and effectively. Finally, an example is given to illustrate the validity and practicability of the model.

Key words: computational power, computational efficiency, computational infrastructure, data center

CLC Number:

WU Meixi, YANG Xiaotong. Five forces of computational power: a comprehensive method to measure computational power[J]. Information and Communications Technology and Policy, 2022, 48(3): 13-21.

Add to citation manager EndNote|Ris|BibTeX

URL:

http://ictp.caict.ac.cn/EN/10.12267/j.issn.2096-5931.2022.03.003

http://ictp.caict.ac.cn/EN/Y2022/V48/I3/13

Figures/Tables 9

References 21

[1]	McMahon FH. The livermore fortran kernels: a computer test of the numerical performance range[R], 1986.
[2]	Sun Y, Agostini NB, Dong Sun, et al. Summarizing CPU and GPU design trends with product data[J]. arXiv preprint arXiv:1911.11313, 2019
[3]	郭亮, 吴美希, 王峰, 等. 数据中心算力评估: 现状与机遇[J]. 信息通信技术与政策, 2021, 47(2):79-86.
[4]	Yue CY. Decision Theory and Method[M]. 北京: 中国科学技术大学出版社, 2003.
[5]	Rao C, Chen W, Wang C. Research and application of fuzzy grey multi-attribute group decision making[C]. 2008 Chinese Control and Decision Conference. IEEE, 2008: 2292-2295.
[6]	Jun G, Tao HU, Ting YU, et al. A multi-attribute group decision-making approach based on uncertain linguistic information[C]. 2019 Chinese Control And Decision Conference (CCDC). IEEE, 2019:4125-4128.
[7]	Fang WG, Hong Z. A multi-attribute group decision-making method approaching to group ideal solution[C]. 2007 IEEE International Conference on Grey Systems and Intelligent Services. IEEE, 2007: 815-819.
[8]	Ye J. Bidirectional projection method for multiple attributegroup decision making with neutrosophic numbers[J]. Neural Computing and Applications, 2015, 28(5):1-9.
[9]	Chen CT. Extensions of the TOPSIS for group decision-making under fuzzy environment[J]. Fuzzy sets and Systems, 2000, 114(1):1-9. doi: 10.1016/S0165-0114(97)00377-1 URL
[10]	Shih HS, Shyur HJ, Lee ES. An extension of TOPSIS for group decision making[J]. Mathematical and Computer Modelling, 2007, 45(7-8):801-813. doi: 10.1016/j.mcm.2006.03.023 URL
[11]	Aiping Y. A novel model for dynamic multi- attributes group decision-making based on vague set and its TOPSIS solution[C]. Proceedings of the 29th Chinese Control Conference. IEEE, 2010:5539-5543.
[12]	张金清. 序方法与均衡分析[M]. 上海: 复旦大学出版社, 2003.
[13]	张小芝, 朱传喜. 多属性决策的广义等级偏好优序法[J]. 系统工程理论与实践, 2013, 33(11):2853-2858.
[14]	Zhang X, Liu P. A grey relational projection method for multi-attribute decision making based on intuitionistic trapezoidal fuzzy number[J]. Applied Mathematical Modelling, 2013, 37(5):3467-3477. doi: 10.1016/j.apm.2012.08.012 URL
[15]	Ye J. Bidirectional projection method for multiple attribute group decision making with neutrosophic numbers[J]. Neural Computing and Applications, 2017, 28(5):1021-1029. doi: 10.1007/s00521-015-2123-5 URL
[16]	Opricovic S, Tzeng GH. Compromise solution by MCDM methods: a comparative analysis of VIKOR and TOPSIS[J]. European Journal of Operational Research, 2004, 156(2):445-455. doi: 10.1016/S0377-2217(03)00020-1 URL
[17]	工业和信息化部. 新型数据中心发展三年行动计划(2021-2023年)[EB/OL].(2021-07-04)[2022-02-01]. http://www.gov.cn/zhengce/zhengceku/2021-07/14/content_5624964.htm.
[18]	Haywood A, Sherbeck J, Phelan P, et al. Investigating a relationship among CPU and system temperatures, thermal power, and CPU tasking levels[C]. 13th InterSociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems. IEEE, 2012: 821-827.
[19]	Nickolls J, Dally W J. The GPU computing era[J]. IEEE micro, 2010, 30(2):56-69. doi: 10.1109/MM.2010.41 URL
[20]	开放数据中心委员会. 数据中心算力白皮书[R], 2021.
[21]	Ignatov A, Timofte R, Denna M, et al. Real-time quantized image super-resolution on mobile npus, mobile ai 2021 challenge: Report[C]. Roceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021: 2525-2534.

编号	通用算力 /TFLOPS	智能算力 /TFLOPS	算效 /GFLOPS/W	网络 /Mbit/s	存储 /IOPS
DC.1	61.25	547.23	72.00	901.00	8.10
DC.2	137.30	545.03	40.30	1699.30	12.70
DC.3	23.90	50.61	23.90	1218.40	3.60
DC.4	20.20	350.32	100.00	964.90	15.40
DC.5	127.07	640.65	80.44	376.30	3.30
DC.6	16.65	88.35	13.24	467.60	4.50

编号	通用算力 /TFLOPS	智能算力 /TFLOPS	算效 /GFLOPS/W	网络 /Mbit/s	存储 /IOPS
DC.1	61.25	547.23	72.00	901.00	8.10
DC.2	137.30	545.03	40.30	1699.30	12.70
DC.3	23.90	50.61	23.90	1218.40	3.60
DC.4	20.20	350.32	100.00	964.90	15.40
DC.5	127.07	640.65	80.44	376.30	3.30
DC.6	16.65	88.35	13.24	467.60	4.50

编号	等级1	等级2	等级3	等级4	等级5
通用算力/TFLOPS	0~10	10~20	20~50	50~100	100~200
智能算力/TFLOPS	0~90	90~130	130~230	230~500	500~1 000
算效/GFLOPS/W	0~10	10~20	20~30	30~50	50~100
网络/Mbit/s	0~300	300~500	500~900	900~1 500	1500~2 200
存储/IOPS	0~1	1~3	3~7	7~12	12~20

编号	等级1	等级2	等级3	等级4	等级5
通用算力/TFLOPS	0~10	10~20	20~50	50~100	100~200
智能算力/TFLOPS	0~90	90~130	130~230	230~500	500~1 000
算效/GFLOPS/W	0~10	10~20	20~30	30~50	50~100
网络/Mbit/s	0~300	300~500	500~900	900~1 500	1500~2 200
存储/IOPS	0~1	1~3	3~7	7~12	12~20

编号	通用算力	智能算力	算效	网络	存储
DC.1	4	5	5	4	4
DC.2	5	5	4	5	5
DC.3	3	1	3	4	3
DC.4	3	4	5	4	5
DC.5	5	5	5	2	3
DC.6	2	1	2	2	3