Power usage effectiveness

Last updated

Power usage effectiveness (PUE) is a ratio that describes how efficiently a computer data center uses energy; specifically, how much energy is used by the computing equipment (in contrast to cooling and other overhead that supports the equipment).

Contents

PUE is the ratio of the total amount of energy used by a computer data center facility [1] [2] [3] [4] [5] [6] [7] [8] [ excessive citations ] to the energy delivered to computing equipment. PUE is the inverse of data center infrastructure efficiency.

PUE was originally developed by a consortium called The Green Grid. PUE was published in 2016 as a global standard under ISO/IEC 30134-2:2016

An ideal PUE is 1.0. Anything that isn't considered a computing device in a data center (e.g. lighting, cooling, etc.) falls into the category of facility energy consumption.

Issues and problems with the power usage effectiveness

The PUE metric is the most popular method of calculating energy efficiency. Although it is the most effective in comparison to other metrics, PUE comes with its share of flaws. This is the most frequently used metric for operators, facility technicians, and building architects to determine how energy efficient their data center buildings are. [9] Some professionals even brag about their PUE being lower than others. Naturally, it is not a surprise that in some cases an operator may “accidentally” not count the energy used for lighting, resulting in lower PUE. This problem is more linked to a human mistake, rather than an issue with the PUE metric system itself.

One real problem is PUE does not account for the climate within the cities the data centers are built. In particular, it does not account for different normal temperatures outside the data center. For example, a data center located in Alaska cannot be effectively compared to a data center in Miami. A colder climate results in a lesser need for a massive cooling system. Cooling systems account for roughly 30 percent of consumed energy in a facility, while the data center equipment accounts for nearly 50 percent. [9] Due to this, the Miami data center may have a final PUE of 1.8 and the data center in Alaska may have a ratio of 1.7, but the Miami data center may be running overall more efficiently. In particular, if it happened to be in Alaska, it may get a better result.

Additionally, according to a case study on Science Direct, "an estimated PUE is practically meaningless unless the IT is working at full capacity". [10]

All in all, finding simple, yet recurring issues such as the problems associated with the effect of varying temperatures in cities and learning how to properly calculate all the facility energy consumption is very essential. By doing so, continuing to reduce these problems ensures that further progress and higher standards are always being pushed to improve the success of the PUE for future data center facilities. [9]

To get precise results from an efficiency calculation, all the data associated with the data center must be included. Even a small mistake can cause many differences in PUE results. One practical problem that is frequently noticed in typically data centers include adding the energy endowment of any alternate energy generation systems (such as wind turbines and solar panels) running in parallel with the data center to the PUE, leading to an obfuscation of the true data center performance. Another problem is that some devices that consume power and are associated with a data center may actually share energy or uses elsewhere, causing a huge error on PUE.

Benefits and limitation

PUE was introduced in 2006 and promoted by The Green Grid (a non-profit organization of IT professionals) in 2007, and has become the most commonly used metric for reporting the energy efficiency of data centres. [10] Although it is named "power usage effectiveness", it actually measures the energy use of the data centre. [10]

The PUE metric has several important benefits. First, the calculation can be repeated over time, allowing a company to view their efficiency changes historically, or during time-limited events like seasonal changes. Second, companies can gauge how more efficient practices (such as powering down idle hardware) affect their overall usage. Finally, the PUE metric creates competition, “driving efficiencies up as advertised PUE values become lower". [10] Companies can then use PUE as a marketing tool.

However, there are some issues with the PUE metric. The main one arises from the way the ratio is calculated. Because IT load is the sole denominator, any reduction in IT load (for example through virtualisation allowing some hardware to be stood down, or simply through more energy-efficient hardware) will cause the PUE to rise. This gives a perverse result.

As well as the issues mentioned in the last paragraph, some other issues are the efficiency of the power supply network and calculating the accurate IT load. According to the sensitivity analysis by Gemma, [10] "Total energy consumption is equal to the total amount of energy used by the equipment and infrastructure in the facility (WT) plus the energy losses due to inefficiencies in the power delivery network (WL), hence: PUE=(WT+WL)/WIT." Based on the equation, the inefficiencies of the power delivery network (WL) will increase the total energy consumption of the data center. The PUE value goes up as the data center becomes less efficient. IT load is another important issue of the PUE metric. "It is crucial that an accurate IT load is used for the PUE, and that it is not based upon the rated power use of the equipment. Accuracy in the IT load is one of the major factors affecting the measurement of the PUE metric, as utilization of the servers has an important effect on IT energy consumption and hence the overall PUE value". [10] For example, a data center with high PUE value and high server utilization could be more efficient than a data center with low PUE value and low server utilization. [10] There is also some concern within the industry of PUE as a marketing tool [11] leading some to use the term "PUE Abuse". [12]

Notably efficient companies

In October 2008, Google's data center was noted to have a ratio of 1.21 PUE across all 6 of its centers, which at the time was considered as close to perfect as possible. Right behind Google was Microsoft, which had another notable PUE ratio of 1.22. [13]

Since 2015, Switch, the developer of SUPERNAP data centers, has had a third party audited colocation PUE of 1.18 for its SUPERNAP 7 Las Vegas, Nevada facility, with an average cold aisle temp of 20.6 °C (69.1 °F) and average humidity of 40.3%. This is attributed to Switch's patented hot aisle containment and HVAC technologies. [14]

As of the end of Q2 2015, Facebook's Prineville data center had a power usage effectiveness (PUE) of 1.078 and its Forest City data center had a PUE of 1.082. [15]

In October 2015, Allied Control has a claimed PUE ratio of 1.02 [16] through the use of two-phase immersion cooling using 3M Novec 7100 fluid.

In January 2016, the Green IT Cube in Darmstadt was dedicated with a 1.07 PUE. [17] It uses cold water cooling through the rack doors.

In February 2017, Supermicro has announced deployment of its disaggregated MicroBlade systems. An unnamed Fortune 100 company has deployed over 30,000 Supermicro MicroBlade servers at its Silicon Valley data center with a (PUE) of 1.06. [18]

Through proprietary innovations in liquid cooling systems, French hosting company OVH has managed to attain a PUE ratio of 1.09 in its data centers in Europe and North America [19] while in 2023 they reported a 12 months overall PUE of 1.29. [20]

In 2021, Google reported a PUE of 1.1 across their data centers worldwide, and less than 1.06 for their best sites. [21] [22]

Standards

PUE was published in 2016 as a global standard under ISO/IEC 30134-2:2016 as well as a European standard under EN 50600-4-2:2016.

See also

Related Research Articles

<span class="mw-page-title-main">Server farm</span> Collection of computer servers

A server farm or server cluster is a collection of computer servers, usually maintained by an organization to supply server functionality far beyond the capability of a single machine. They often consist of thousands of computers which require a large amount of power to run and to keep cool. At the optimum performance level, a server farm has enormous financial and environmental costs. They often include backup servers that can take over the functions of primary servers that may fail. Server farms are typically collocated with the network switches and/or routers that enable communication between different parts of the cluster and the cluster's users. Server "farmers" typically mount computers, routers, power supplies and related electronics on 19-inch racks in a server room or data center.

<span class="mw-page-title-main">Data center</span> Building or room used to house computer servers and related equipment

A data center or data centre is a building, a dedicated space within a building, or a group of buildings used to house computer systems and associated components, such as telecommunications and storage systems.

Green computing, green IT, or ICT sustainability, is the study and practice of environmentally sustainable computing or IT.

<span class="mw-page-title-main">Urs Hölzle</span> Swiss computer scientist

Urs Hölzle is a Swiss software engineer and technology executive. As Google's eighth employee and its first VP of Engineering, he has shaped much of Google's development processes and infrastructure, as well as its engineering culture. His most notable contributions include leading the development of fundamental cloud infrastructure such as energy-efficient data centers, distributed compute and storage systems, and software-defined networking. Until July 2023, he was the Senior Vice President of Technical Infrastructure and Google Fellow at Google. In July 2023, he transitioned to being a Google Fellow only.

The Climate Savers Computing Initiative was a nonprofit group of consumers, businesses and conservation organizations dedicated to promoting smart technologies that improve power efficiency and reduce energy consumption of computers. Formed in 2007, it was based in Portland, Oregon. In July 2012, Climate Savers Computing Initiative combined with The Green Grid and its programs continue within that organization.

In computing, performance per watt is a measure of the energy efficiency of a particular computer architecture or computer hardware. Literally, it measures the rate of computation that can be delivered by a computer for every watt of power consumed. This rate is typically measured by performance on the LINPACK benchmark when trying to compare between computing systems: an example using this is the Green500 list of supercomputers. Performance per watt has been suggested to be a more sustainable measure of computing than Moore's Law.

The Green Grid is a nonprofit, industry consortium of end-users, policy-makers, technology providers, facility architects, and utility companies collaborating to improve the resource efficiency of data centers.

Miscellaneous electric loads (MELs) in buildings are electric loads resulting from a multitude of devices (electronic and other) excluding main systems for space heating, cooling, water heating, or lighting. MELs are produced by hard-wired and “plug-in” electrical devices that draw power, including office equipment such as desktop computers and monitors, mobile electronics (laptops, tablets, mobile phones, and their charging units), printers, fans, task lighting, and home equipment such as home entertainment centers, kitchen electronics (microwaves, toaster ovens, cooking accessories), bath items (hair dryers, lighted mirrors, and electric hot tubs), and other devices such as security systems and ceiling fans. MELs are gaining greater importance in energy management as personal electronics proliferate and become standard across demographic groups. MELs demand has been rising as a percentage of total energy end-use and is expected to continue rising.

Data center infrastructure efficiency (DCIE), is a performance improvement metric used to calculate the energy efficiency of a data center. DCIE is the percentage value derived, by dividing information technology equipment power by total facility power.

IT energy management or Green IT is the analysis and management of energy demand within the Information Technology department in any organization. IT energy demand accounts for approximately 2% of global CO2 emissions, approximately the same level as aviation, and represents over 10% of all the global energy consumption. IT can account for 25% of a modern office building's energy cost.

An Energy Rebate Program, or Energy Credit Incentive Program, provides a cash rebate program for customers planning to install new, energy efficient information technology (IT) equipment or cooling systems. These programs push companies to construct more energy efficient data centers, or to consolidate compute, storage and networking resources via virtualization technologies.

Green Power Usage Effectiveness (GPUE) is a proposed measurement of both how much sustainable energy a computer data center uses, its carbon footprint per usable kilowatt hour (kWh) and it uses its power; specifically, how much of the power is actually used by the computing equipment. It is an addition to the power usage effectiveness (PUE) definition and was first proposed by Greenqloud.

<span class="mw-page-title-main">HP Performance Optimized Datacenter</span> Portable data centre

The HP Performance Optimized Datacenter (POD) is a range of three modular data centers manufactured by HP.

NABERS, the National Australian Built Environment Rating System, is an initiative by the government of Australia to measure and compare the environmental performance of Australian buildings and tenancies. There are NABERS rating tools for commercial office buildings to measure greenhouse gas emissions, energy efficiency, water efficiency, waste efficiency and indoor environment quality. There are also energy/greenhouse and water rating tools for hotels, shopping centres and data centres.

Energy Logic is a vendor-neutral approach to achieving energy efficiency in data centers. Developed and initially released in 2007, the Energy Logic efficiency model suggests ten holistic actions – encompassing IT equipment as well as traditional data center infrastructure – guided by the principles dictated by the "Cascade Effect."

Water Usage Effectiveness (WUE) is a sustainability metric created by The Green Grid in 2011 to attempt to measure the amount of water used by datacenters to cool their IT assets. To calculate simple WUE, a data center manager divides the annual site water usage in liters by the IT equipment energy usage in kilowatt hours (Kwh). Water usage includes water used for cooling, regulating humidity and producing electricity on-site. More complex WUE calculations are available from The Green Grid website.

<span class="mw-page-title-main">Immersion cooling</span> IT cooling practice

Immersion cooling is an IT cooling practice by which complete servers are immersed in a dielectric, electrically non-conductive fluid that has significantly higher thermal conductivity than air. Heat is removed from a system by putting the coolant in direct contact with hot components, and circulating the heated liquid through heat exchangers. This practice is highly effective because liquid coolants can absorb more heat from the system, and are more easily circulated through the system, than air. Immersion cooling has many benefits, including but not limited to: sustainability, performance, reliability and cost

<span class="mw-page-title-main">Supermicro</span> American supplier of servers and other information technology products

Super Micro Computer, Inc., dba Supermicro, is an American information technology company based in San Jose, California. It has manufacturing operations in the Silicon Valley, the Netherlands and at its Science and Technology Park in Taiwan. Founded on November 1, 1993, Supermicro is one of the largest producers of high-performance and high-efficiency servers. It also provides server management software, and storage systems for various markets, including enterprise data centers, cloud computing, artificial intelligence, 5G and edge computing.

The Zettabyte Era or Zettabyte Zone is a period of human and computer science history that started in the mid-2010s. The precise starting date depends on whether it is defined as when the global IP traffic first exceeded one zettabyte, which happened in 2016, or when the amount of digital data in the world first exceeded a zettabyte, which happened in 2012. A zettabyte is a multiple of the unit byte that measures digital storage, and it is equivalent to 1,000,000,000,000,000,000,000 (1021) bytes.

<span class="mw-page-title-main">Green data center</span> Server facility which utilizes energy-efficient technologies

A green data center, or sustainable data center, is a service facility which utilizes energy-efficient technologies. They do not contain obsolete systems, and take advantage of newer, more efficient technologies.

References

  1. "What is power usage effectiveness (PUE)? - Definition from WhatIs.com".
  2. Digital Realty - PUE Data Center Efficiency Metric
  3. "Optimizing Power Usage Effectiveness In Data Centers". www.esmagazine.com.
  4. "The Green Grid - The Green Grid Data Center Power Efficiency Metrics: PUE and DCiE".
  5. Google - Efficient Computing - Data Centers - Efficiency: How we do it?
  6. Dell - Best Practices for Increasing Data Center Energy Efficiency
  7. Cisco Systems - Cisco Energywise
  8. Google Data Center Optimization -
  9. 1 2 3 Yuventi, Jumie; Mehdizadeh, Roshan (2013). "A critical analysis of Power Usage Effectiveness and its use in communicating data center energy consumption". Energy and Buildings. 64: 90–94. doi:10.1016/j.enbuild.2013.04.015.
  10. 1 2 3 4 5 6 7 Brady, Gemma, Nikil Kapur, Jonathan Summers, and Harvey Thompson. "A Case Study and Critical Assessment in Calculating Power Usage Effectiveness for a Data Centre." Energy Conversion & Management, 76 (2013): 155-161.
  11. "Romonet website, Blogs, 'Look at the size of my PUE!'". Archived from the original on 2015-09-24. Retrieved 2015-09-04.
  12. "PUE abuse beyond the pale? - Datacenter Dynamics". archive.datacenterdynamics.com.
  13. Tuf, Steve. "Power Usage Effectiveness." It.toolbox. Toolbox, n.d. Web. 17 Nov. 2014.
  14. Miller, Rich. “Inside SUPERNAP 8: Switch’s Tier IV Data Fortress.” Data Center Knowledge. Feb. 11, 2014
  15. "Energy Efficiency". Open Compute Project . Open Compute Project. Archived from the original on 12 March 2016. Retrieved 18 March 2016.
  16. "Two-Phase Immersion Cooling" http://multimedia.3m.com/mws/media/1127920O/2-phase-immersion-coolinga-revolution-in-data-center-efficiency.pdf
  17. "Green IT Cube: Hocheffizientes Supercomputer-Domizil eingeweiht". 2016-01-23. Retrieved 2016-01-24.
  18. "Supermicro - News - Supermicro Deploys 30,000+ MicroBlade™ Servers to Enable One of the World's Highest Efficiency (1.06 PUE) Data Centers". www.supermicro.com.
  19. OVH. "A green hosting provider - OVH Canada". www.ovh.com.
  20. "Managing our environmental impact every step of the way". ovhcloud.com.
  21. "Google Data Centers efficiency". google.com.
  22. "How Google plans to use 100% carbon-free energy in its data centers by 2030". cnbc.com.