Top Banner
How green is the internet? 1
31

Koomey on Internet infrastructure energy 101

Jul 04, 2015

Download

Technology

Jonathan Koomey

This talk, given at Google on June 6, 2012, summarizes what we know about energy use and information technology in a clear and understandable way. The person preceding me on stage was former Vice President Al Gore, so the pressure was on! I think I delivered, but you be the judge.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Koomey on Internet infrastructure energy 101

How green is the internet?

1  

Page 2: Koomey on Internet infrastructure energy 101

Internet Infrastructure Energy 101 Jonathan Koomey, Ph.D.

Research Fellow, Steyer-Taylor Center for Energy Policy and Finance

Stanford University http://www.koomey.com

June 6, 2013

2  

Page 3: Koomey on Internet infrastructure energy 101

Defining “the Internet”

3  

Page 4: Koomey on Internet infrastructure energy 101

The big picture view

Source: Ericsson and TeliaSonera (Malmodin and Lundén et al 2013) with support from CESC, KTH Sweden

Key components

•  Data centers •  Core network •  Access networks •  End-user

communications equipment

•  End-user computing equipment

Lots of complexity here! 4  

Page 5: Koomey on Internet infrastructure energy 101

The Internet is data…

Total data flows

1986

1993

2000

2007

Source: Adapted from Hilbert et. al. 2011

Mobile data

Fixed Internet

Voice

5  

Page 6: Koomey on Internet infrastructure energy 101

…but it’s also physical

Photo sources (clockwise from top left): Google. Flickr users Mr. T in DC, digger_90_tristar, geerlingguy, alachia, antonionicolaspina 6  

Page 7: Koomey on Internet infrastructure energy 101

0 500 1000 1500 2000

IP core network

Operator activities

Access networks

Other user equipment

Data centers and LANs (3rd party)

User PCs

Electricity use, GWh/year

What matters most….

These are key

Source: Ericsson and TeliaSonera (Malmodin and Lundén et al 2013) with support from CESC, KTH Sweden. Data are for Sweden, circa 2010.

7  

Page 8: Koomey on Internet infrastructure energy 101

The big three

End-user equipment

Data centers

Access networks

8  

Page 9: Koomey on Internet infrastructure energy 101

End-user equipment

9  

Page 10: Koomey on Internet infrastructure energy 101

End-user equipment Computing

– Desktops and local servers – Laptops – Tablets

Communications – Phones – Wireless routers – Set-top boxes – Switches

Display – Computer monitors – TVs (IP connected)

Ultra low-power computing/sensors (small but growing) Photo sources (clockwise from top left): Flickr users sucello, expertcomp, verdammtescheissenochmal, janitors

10  

Page 11: Koomey on Internet infrastructure energy 101

Growing base of devices worldwide

Desktops and laptops Servers

Source: IDC 2013 Vernon Turner 11  

Page 12: Koomey on Internet infrastructure energy 101

A key computing trend…

Source: Koomey et. al. 2011

Mobile systems and sensors becoming widespread, driven by progress in computing efficiency (100x every decade)

12  

Page 13: Koomey on Internet infrastructure energy 101

…led to the rise of tablets and mobile phones

Source: IDC (http://www.idc.com/getdoc.jsp?containerId=prUS24129713)

Tablet shipments = desktops in 2012!

Source: Hilbert and López 2012a and 2012b

13  

Page 14: Koomey on Internet infrastructure energy 101

Embedded emissions from manufacturing

Source: Koomey et. al. 2013

0% 20% 40% 60% 80% 100%

Server (Mac Mini OS X server)

Laptop computer (Macbook Pro 13")

Smart Phone (iPhone5)

NAND Flash memory - 1 GB

Share of CO2 emissions

Production Operation

0 200 400 600 800 1000 1200 Life Cycle CO2 emissions (kg)

Percentage contributions Absolute emissions

14  

Page 15: Koomey on Internet infrastructure energy 101

Data centers

15  

Page 16: Koomey on Internet infrastructure energy 101

Data center electricity use worldwide

Source: Koomey 2011. Graph shows worldwide numbers. For the US, the range for data centers in 2010 was 1.7 to 2.2% of the total. N.B. Infrastructure in this slide refers to cooling, fans, pumps, and power distribution inside data centers. 16  

Page 17: Koomey on Internet infrastructure energy 101

Data center lessons

Big inefficiencies in “in-house” data centers (cloud providers much better) Just adopting best practices will save 50+% Biggest impediments to efficiency are institutional, not technical IT efficiency most important, followed by infrastructure efficiency and sourcing of low-carbon electricity

17  

Page 18: Koomey on Internet infrastructure energy 101

Access networks

18  

Page 19: Koomey on Internet infrastructure energy 101

Access network bandwidth installed worldwide in 2010

Source: Hilbert and López 2012a and 2012b

- 500 1,000 1,500 2,000

2G mobile data Fixed line phone All mobile voice

Other 3G mobile data

2.5G mobile data Fiber

Cable Modem DSL

Installed capacity TB/second

0

19  

Page 20: Koomey on Internet infrastructure energy 101

Source: Ericsson and TeliaSonera (Malmodin and Lundén et al 2013) with support from CESC, KTH Sweden.

0 50 100 150 200

Fixed cable-TV and fiber broadband

Fixed xDSL broadband

3G (WCDMA)

PSTN and VoIP

2G (GSM)

Electricity use, GWh/year

Access network electricity use (Sweden 2010)

20  

Page 21: Koomey on Internet infrastructure energy 101

System effects of IT

21  

Page 22: Koomey on Internet infrastructure energy 101

System effects of IT

Dematerialization (move bits, not atoms) – CDs vs. downloads

Big-systems optimization

– Smart parking sensors reduce traffic Enabling structural change

– Flatter, more nimble organizations

22  

Page 23: Koomey on Internet infrastructure energy 101

Dematerialization: Move bits not atoms

Source: Weber et. al. 2010

CO2 emissions for downloads and physical CDs

-80% -40%

23  

Page 24: Koomey on Internet infrastructure energy 101

Big systems optimization: Smart parking

Source: Mark Noworolski, Streetline Networks

Motes use <400µW on Average. For LA, with 40,000 parking spots, that implies total mote power of about 15W. Mote technology is from Dust Networks

24  

Page 25: Koomey on Internet infrastructure energy 101

IT enables “business process redesign”, improving efficiency across the board. Example:

– Gave suppliers access to POS and inventory data, as well as company forecasts

– Pioneered aggressive use of RFID – Improved the flow of supplies and finished goods – The result: better coordination of suppliers with Walmart’s needs, plus much lower distribution costs

Structural change

For details on the Walmart example, see Traub 2012. For more examples, see Brynjolfsson and Hitt 2000. 25  

Page 26: Koomey on Internet infrastructure energy 101

Key research issues

26  

Page 27: Koomey on Internet infrastructure energy 101

Key research issues

Need recent data on electricity use and potential savings Need more and better automated reporting of

– Energy use – User behavior

Average (fixed) vs. marginal (variable) energy use

– Most devices have high fixed energy use – Be careful to distinguish average vs. marginal effects

Need more system efficiency case studies 27  

Page 28: Koomey on Internet infrastructure energy 101

Conclusions

28  

Page 29: Koomey on Internet infrastructure energy 101

Conclusions

Popular preoccupation with electricity used by Internet-related systems is misplaced

– Probably < 10% of total electricity but not well characterized – End-user devices important, but most can’t be clearly allocated to “the Internet”

System effects potentially much more important than direct electricity use

– IT affects efficiency in the other 90% of electricity use plus all the fuels

Updated data needed! 29  

Page 30: Koomey on Internet infrastructure energy 101

References

30  

Page 31: Koomey on Internet infrastructure energy 101

References Brynjolfsson, Erik, and Lorin M. Hitt. 2000. "Beyond Computation: Information Technology, Organizational Transformation and Business Performance." Journal of Economic Perspectives. vol. 14, no. 4. Fall. pp. 23-48.

Hilbert, Martin, and Priscila López. 2011. "The World's Technological Capacity to Store, Communicate, and Compute Information." Science. vol. 332, no. 6025. April 1. pp. 60-65.

Hilbert, Martin, and Priscila López. 2012a. "Info Capacity| How to Measure the World’s Technological Capacity to Communicate, Store and Compute Information? Part I: Results and Scope." International Journal of Communication. vol. 6, pp. 956-979. [http://ijoc.org/ojs/index.php/ijoc/article/view/1562/742]

Hilbert, Martin, and Priscila López. 2012b. "Info Capacity| How to Measure the World’s Technological Capacity to Communicate, Store and Compute Information? Part II: Measurement Unit and Conclusions." International Journal of Communication. vol. 6, pp. 936-955. [http://ijoc.org/ojs/index.php/ijoc/article/view/1563/741]

Koomey, Jonathan. 2008. "Worldwide electricity used in data centers." Environmental Research Letters. vol. 3, no. 034008. September 23. <http://stacks.iop.org/1748-9326/3/034008>.

Koomey, Jonathan G., Stephen Berard, Marla Sanchez, and Henry Wong. 2011. "Implications of Historical Trends in The Electrical Efficiency of Computing." IEEE Annals of the History of Computing. vol. 33, no. 3. July-September. pp. 2-10. http://www.computer.org/csdl/mags/an/2011/03/man2011030046-abs.html

Koomey, Jonathan. 2011. Growth in data center electricity use 2005 to 2010. Oakland, CA: Analytics Press. August 1. http://www.analyticspress.com/datacenters.html

Koomey, Jonathan G., H. Scott Matthews, and Eric Williams. 2013. "Smart Everything: Will Intelligent Systems Reduce Resource Use?" In press at The Annual Review of Environment and Resources. May.

Masanet, Eric, Arman Shehabi, and Jonathan Koomey. 2013. "Characteristics of Low-Carbon Data Centers." In press at Nature Climate Change. May.

Malmodin, Jens, Dag Lundén, Åsa Moberg, Greger Andersson, and Mikael Nilsson. 2013. "Life cycle assessment of ICT networks–carbon footprint and operational electricity use from the operator, national and subscriber perspective." Submitted to The Journal of Industrial Ecology. March 8.

Traub, Todd. 2012. "Wal-mart used technology to become supply chain leader." In Arkansas Business. July 2. [http://www.arkansasbusiness.com/article/85508/wal-mart-used-technology-to-become-supply-chain-leader]

Weber, Christopher, Jonathan G. Koomey, and Scott Matthews. 2010. "The Energy and Climate Change Impacts of Different Music Delivery Methods." The Journal of Industrial Ecology. vol. 14, no. 5. October. pp. 754–769. [http://dx.doi.org/10.1111/j.1530-9290.2010.00269.x]

31