Research on evaluation of computing power and efficiency in data center:status and opportunities

doi:10.12267/j.issn.2096-5931.2021.02.014

Abstract

Abstract: This paper reviews the research status of computing power and efficiency, including SPEC CPU, SPEC power, MLperf, etc. , which describe the computing power of a single server in the data center, while TOP 500 and green 500 describe the supercomputing power, while PUE describes the power utilization efficiency. And there is no scientific methods to describe the computing power and efficiency of the data center. Based on the analysis, this paper puts forward a method to measure the computing power and efficiency of data centers, and calculates the current level of computing power and efficiency of data centers in China.

Key words: data center, computational power, computational efficiency

GUO Liang, WU Meixi, WANG Feng, GONG Min. Research on evaluation of computing power and efficiency in data center:status and opportunities[J]. Information and Communications Technology and Policy, 2021, 47(2): 79-86.

Add to citation manager EndNote|Ris|BibTeX

URL:

http://ictp.caict.ac.cn/EN/10.12267/j.issn.2096-5931.2021.02.014

http://ictp.caict.ac.cn/EN/Y2021/V47/I2/79

References 25

[1]	国家发展和改革委员会. 国家发改委举行4 月份例行新闻发布会[EB / OL] . (2020-04-20) [2020-12-29] .
	http:/ / www. scio. gov. cn/ xwfbh/ gbwxwfbh/ xwfbh/ fzggw/Document / 1677563/ 1677563. htm.
[2]	郭亮. 边缘数据中心关键技术和发展趋势[J]. 信息通信技术与政策, 2019(12):55-58.
[3]	Top500. TOP500 November 2020[EB / OL]. (2020-11-25) [2020-12-29] . https:/ / www. top500. org/ lists/ top500/2020/ 11/ .
[4]	Top500. GREEN500 November 2020[EB / OL]. (2020-11-25) [ 2020-12-29] . https:/ / www. top500. org / lists /green500/ 2020/ 11.
[5]	Standard Performance Evaluation Corporation[EB / OL].(2020-12-15)[2020-12-29]. http:/ / www. spec. org.
[6]	Standard Performance Evaluation Corporation. SPECCPU® 2017 [ EB / OL] . (2020-11-25) [2020-12-29] .http:/ / www. spec. org / cpu2017/ .
[7]	Standard Performance Evaluation Corporation. SPECPower®[EB / OL]. (2020-12-16)[2020-12-29]. http:/ /www. spec. org / power_ssj2008/ .
[8]	Standard Performance Evaluation Corporation. SPECSERT® Suite [ EB / OL ] . ( 2020-11) [ 2020-12-29] .
	http:/ / www. spec. org / sert / .
[9]	MLPerf. MLPerf Inference[EB/ OL]. (2020-08-25)[2020-12-29]. https:/ / www. mlperf. org/ inference-overview/ .
[10]	MLCommons. October 21, 2020—Inference: Datacenterv0. 7 Results [ EB / OL ] . ( 2020-12) [ 2020-12-29] .
	https:/ / mlcommons. org / en / inference-datacenter-07/ .
[11]	MLCommons. 07. 29. 2020—Mountain View, CAMLPerf Training v0.7 results [EB/ OL]. (2020-12)[2020-12-29].https:/ / mlcommons. org/ en/ news/ mlperf-training-v07/ .
[12]	开放数据中心委员会[EB / OL]. (2020-12)[2020-12-29]. http:/ / www. odcc. org. cn / index. html.
[13]	开放数据中心委员会. 服务器能效评测规范[EB /OL] . (2020-12) [2020-12-29] . http:/ / www. odcc.org. cn / download / p-1169551993955831810. html.
[14]	The Green Grid. WP # 49 PUETM : a comprehensive examination of the metric[R], 2011.
[15]	John Nickolls, William J Dally. The gpu computing era[J]. IEEE micro, 2010,30(2):56-69.
[16]	McMahon, F H. The livermore fortran kernels: a computer test of the numerical performance range [J] .
	United States: N, 1986, 179.
[17]	Yifan Sun. Summarizing CPU and GPU design trends with product data[Z], 2019.
[18]	Intel. Intel product specifications[EB / OL]. (2020-12)[2020-12-29]. https:/ / ark. intel. com / .
[19]	Nvidia. 适用于服务器的TESLA 数据中心GPU[EB /OL]. (2020-12)[2020-12-29]. https:/ / www. nvidia.
	cn / data-center / tesla / .
[20]	郭亮. 面向云计算的企业级硬盘基准测试[J]. 电信网技术, 2016(10):1-4.
[21]	OpenAI. AI and Compute[EB/ OL]. (2019-11-07)[2020-12-29]. https://openai. com/ blog / ai-and-compute / .
[22]	中国信息通信研究院, 开放数据中心委员会. 数据中心白皮书(2020)[R], 2020.
[23]	工业和信息化部信息通信发展司. 全国数据中心应用发展指引(2019)[M]. 人民邮电出版社, 2020.
[24]	Gartner. Servers marketshare WW country 2019Q4[R],2019.
[25]	IDC. Datacenter processing, 4Q19: server CPU, GPU,