【新年重磅part-2】全球数据中心能耗评估再校准
更多干货等你发现！
全球数据中心能耗评估再校准-2
Recalibrating global data center energy-use estimates - 2
本文为该系列的第二篇，首篇回顾：全球数据中心能耗评估再校准-1
随着全球数据中心能效的持续改善，未来3到4年，尽管数据中心的计算力将翻一番，但是相对应的数据中心能耗却基本不变。但这之后数据中心能耗将迎来更大挑战。
随着对数据中心需求的增加，IT设备和相应冷却系统的能效改善可使能耗得到控制
As demand for data centers rises, energy efficiency improvements to the IT devices and cooling systems they house can keep energy use in check
与以前的数据相比，这里利用的数据更强调了技术导向和时间的一致性。自2011年以来，思科的分析师发布了有关全球服务器库存、数据中心工作负载、服务器虚拟化水平以及传统、云和最新超大规模数据中心存储容量的数据和展望[1]。从2016年开始的一系列报告中，劳伦斯·伯克利国家实验室发布了数据中心中常用服务器、存储设备和网络设备的能源趋势分析[8、11、13]。分析人员记录了超大规模数据中心的数量和位置，而这些数据中心占了全球数据中心计算实例的很大一部分，而主要的数据中心运营商也开始更多地报告其PUE值[14]。
The data leveraged here facilitate a more technology-richand temporally consistent approach than was available previously. Since 2011,analysts at Cisco have published data and outlooks for worldwide server stocks,data center workloads, server virtualization levels, and storage estimates fortraditional, cloud, and, most recently, hyperscale data centers (1). In aseries of reports starting in 2016, Lawrence Berkeley National Laboratory haspublished energy trend analyses of servers, storage devices, and networkdevices commonly used within data centers (8, 11, 13). Analysts have documentedthe numbers and locations of hyperscale data centers that represent asubstantial fraction of global data center compute instances, and major datacenter operators are increasingly reporting their PUE (14).
将这些数据集成到一个自下而上的模型框架中后，结果表明，尽管自2010年以来全球数据中心的能耗略有增加，但能耗的增长与同期数据中心计算实例的增长已基本解耦（见图2的第2部分）。此外，这些新数据提供的精确视图表明，2010年全球数据中心的能耗约为194 TWh，略低于2010年自下而上研究的评估值下限（203 TWh），而当时可用数据较少[9]。
When integrated into a bottom-up modeling frame-work,these data suggest that, although global data center energy has increasedslightly since 2010, growth in energy use has been substantially de-coupledfrom growth in data center compute instances over the same time period (see thesecond figure, second graph). Moreover, the refined view provided by these newdata suggests that global data center energy use in 2010 was around 194 TWh,slightly less than the lower-bound estimate in the 2010 bottom-up study (203TWh) when fewer data were available[9].
2018年，我们估算的全球数据中心能耗上升至205TWh，约占全球用电量的1％。与2010年相比，增长了6％，而同期全球数据中心计算实例增长了550％。以每个计算实例的能耗表示，自2010年以来，全球数据中心的能耗强度每年下降20％，与其他主要需求部门（例如航空和工业）的年度能效增长相比有了显著改善，后者的能效改善则低了一个数量级[12]。
In 2018, we estimated that global data center energy userose to 205 TWh, or around 1% of global electricity consumption. Thisrepresents a 6% increase compared with 2010, whereas global data center computeinstances increased by 550% over the same time period. Expressed as energy useper compute instance, the energy intensity of global data centers has decreasedby 20% annually since 2010, a notable improvement compared with recent annualefficiency gains in other major demand sectors (e.g., aviation and industry),which are an order of magnitude lower (12).
新的综合数据阐明了一些关键的技术和结构趋势，有助于解释这些大幅度的能源强度改善（见图2第2部分）。服务器效率提高和服务器虚拟化（减少了每个计算实例所需的服务器耗电量）的结合使计算实例增加了6倍，而全球服务器能耗仅增加了25％，存储驱动效率和密度的提高使存储容量增加了25倍，而全球存储能耗仅增加了3倍。向更快、更节能的端口技术转变已使数据中心IP流量增加了10倍，而网络设备的能耗仅略有增加。总而言之，尽管IT设备（服务器，存储和网络）的总体能耗已从2010年的约92 TWh增加到2018年的约130 TWh，技术和运营效率的提高使服务获得了大幅增长，而能耗增量却相对较小。
The new integrated data illuminate some key technologicaland structural trends that help explain these large energy intensityimprovements (see the first figure and the second figure, second graph). Thecombination of increased server efficiencies and greater server virtualization(which reduces the amount of server power required for each compute instance)has enabled a six-fold increase in compute instances with only a 25% increasein global server energy use, whereas the combination of increased storage-driveefficiencies and densities has enabled a 25-fold increase in storage capacitywith only a threefold increase in global storage energy use. Shifts to fasterand more energy-efficient port technologies have enabled a 10-fold increase indata center IP traffic with only modest increases in net-work device energyuse. In sum, although overall energy use of IT devices (servers, storage, andnetwork) has increased from around 92 TWh in 2010 to around 130 TWh in 2018,technological and operational efficiency gains have enabled substantial growthin services with comparatively much smaller growth in energy use.
值得注意的是，新数据还表明数据中心基础设施系统的能耗大幅减少（如制冷和电源供应），足以抵消大部分IT设备总能耗的增长。基础设施系统能耗的不断减少可以通过服务器的迁移解释，这些服务器正在从较小的传统数据中心（2010年占计算实例的79％）向更大、更节能的云（包括超大规模）数据中心（2018年占计算实例的89％）迁移（见图2第3部分），后座有先进的冷却系统和电源效率，其报告的PUE值要低得多[1，11]。
Notably, the new data also suggest a large decrease inthe energy use of data center infrastructure systems (i.e., cooling and powerprovisioning), enough to mostly offset the growth in total IT device energy use.This decrease is explainable by ongoing shifts in servers away from smallertraditional data centers (79% of compute instances in 2010) and toward largerand more energy-efficient cloud (including hyperscale) data centers (89% ofcompute instances in 2018) (see the second figure, third graph), which havemuch lower reported PUE values owing to cutting-edge cooling-system andpower-supply efficiencies (1, 11).
但是，鉴于对数据中心服务的需求不断增长，当前的能效趋势还能持续多久？众所周知，预测IT设备的长期效率极限众所周知非常困难，尤其是考虑到潜在的颠覆性技术（如量子计算），其能耗尚不清楚[2]。然而，在短期内，市场分析家预测，更高级别的服务器虚拟化是可行的[1]，技术研究表明，IT设备效率仍有提升的潜力空间，包括更多转向低功耗存储设备（8）。在基础设施方面，世界一流的超大规模数据中心已经在以1.1或更低的PUE值运行，接近实际最小值。预计短期内小型传统数据中心还将向超大规模数据中心迁移[1]，这表明基础设施的能耗增长可能会进一步受到抑制。如果这些趋势在未来几年内得以体现，我们的方法表明，将有足够的能效资源来吸收数据中心计算实例的下一个翻番，同时全球数据中心的能耗增量却可以忽略不计（见图2第2部分）。
Yet given ever-growing demand for data center services,how much longer can these current efficiency trends last? Predicting thelong-term efficiency limits of IT devices is notoriously difficult, especiallyin light of potential game-changing technologies such as quantum computing, forwhich energy use is unclear (2). Yet over the near term, market analystspredict that even greater levels of server virtualization are feasible (1), andtechnology studies indicate remaining potential for IT device efficiency gains,including more shifts to low-power storage devices (8). On the infrastructureside, world-class hyperscale data centers are already operating with PUEs of1.1 or lower, which is close to the practical minimum value. Additionalstructural shifts from smaller traditional data centers to hyperscale datacenters are predicted in the near term (1), indicating that infrastructureenergy use may be dampened even further. Should these trends play out over thenext few years, our approach indicates that there is a sufficient energyefficiency resource to absorb the next doubling of data center computeinstances that would occur in parallel with a negligible increase in globaldata center energy use (see the second figure, second graph).
这些调查结果与能源需求必定快速增长的近期预测相反。然而，IT行业、数据中心运营商和决策者不能止步不前。一旦现有的能效资源被完全用尽，将需要努力应对可能急剧增长的能源需求。全球数据中心计算实例的下一个翻番可能会在接下来的3-4年内发生[1]。
These findings lie in contrast to recent predictions ofrapid and unavoidable near-term energy demand growth. Yet the IT industry, datacenter operators, and policy-makers can’t rest on their laurels; diligentefforts will be required to manage possibly sharp energy demand growth once theexisting efficiency resource is fully tapped. The next doubling of global datacenter compute instances may occur within the next 3 to 4 years [1]。
中文翻译：何海
原文出处：POLICY FORUM【Recalibrating global data center energy-use estimates】
REFERENCES AND NOTES 参考文献
1. Cisco, “Cisco Global Cloud Index: Forecast and methodology, 2016–2021 white paper” (Cisco, document 1513879861264127 , 2018).
2. International Energy Agency (IEA), Digitalization & Energy(IEA, 2017).
3. L. Belkhir, A. Elmeligi, J. Clean. Prod.177, 448 (2018).
4. A.S.G. Andrae, T. Edler, Challenges6, 117 (2015).
5. T . Bawdy, “Global warming: Data centres to consume three times as much energy in next decade, experts warn, ” The Independent, 23 January 2016.
6. N. Jones, Nature561, 163 (2018).
7. E. Masanet, R. E. Brown, A. Shehabi, J. G. Koomey,
B. Nordman, Proc. IEEE99, 1440 (2011).
8. A. Shehabi et al., “United States data center energy usage
report” (Lawrence Berkeley National Laboratory, LBNL-
1005775, 2016).
9. J. G. Koomey, “Growth in data center electricity use 2005
to 2010” (Analytics Press for the New Y ork Times, 2011).
10. B. Wagner, “Intergenerational energy efficiency of Dell
EMC PowerEdge servers” (Dell, DellEMC white paper,
2018).
11. A. Shehabi, S. J. Smith, E. Masanet, J. Koomey, Environ. Res.
Lett.13, 124030 (2018).
12. IEA, “T racking clean energy progress” (IEA, 2019);
www.iea.org/tcep/.
13. H. Fuchs et al., Energy Effic. 10.1007/s12053-019-09809-8 (2019).
14. M. Avgerinou, P. Bertoldi, L. Castellazzi, Energies10, 1470 (2017).
15. E. Masanet, A. Shehabi, J. G. Koomey, Nat. Clim. Chang.3, 627 (2013).