WHEB Commentary

Tim Dieppe

The Power of Data


Earlier this year we reached the point where 90% of all the digital data in the world was created in the last two years. This extraordinary statistic almost makes it sound as if we were living in the dark ages back in 2010 when we knew so little! Cisco forecasts that global IP traffic will grow at a compound rate of 23% from 2012 to 2017. YouTube now has 100 hours of video uploaded to its site every minute, while Cisco tell us that they expect a million minutes of video to cross the internet every second in 2017.

The days of the BBC overwriting tapes so that some episodes of Dr Who are lost forever are long gone. All this data is stored for posterity – and most of it is stored in power hungry data centres. We often do not appreciate how much power is required to store all this digital data. Power is by far and away the biggest cost item for a data centre. On average, one data centre uses the equivalent power of 25,000 homes in the western world. This requires significant planning and enabling of grid connectivity to achieve. McKinsey have estimated that the worldwide carbon emissions from data centres will quadruple by 2020, exceeding emissions of the airline industry and thus making a significant contribution to climate change. It is in all of our interests, therefore, to ensure that this data is stored in as efficient and low carbon a manner as possible, and this is where the pure play external data centre providers are able to help.

The data centre industry has become very focussed on power efficiency and has developed several measures to assess this. The first is Power Usage Effectiveness (PUE) which is the ratio of power consumed by the data centre to the power consumed by the actual IT equipment. A perfect PUE would be 1.0, but that would mean that no power at all is being used for overheads and cooling equipment. The Uptime Institute industry survey 2013 reported average PUE of 1.65 and the best providers are known to do much better than this. Facebook, for example provide real time data on their data centre in Oregon, and display PUEs around 1.1.  This is clearly vastly more efficient than using internal data servers with all the associated cooling equipment and demonstrates the clear efficiency benefit from using dedicated focussed data centres.

PUE can then be taken a step further to CUE or Carbon Usage Effectiveness which is the CO2 emitted per unit of energy multiplied by the PUE. This encourages both efficient energy use and the use of low carbon or renewable energy, with the ideal CUE being 0.0. Telecity reported CUE of 0.65 with clear targets to reduce CUE by 3% pa. Interxion say that approximately 75% of their power comes from sustainable sources and that they have improved PUE every year since 2003.

It is clear, then, that the data centre companies are able to host data in a much more efficient and low carbon manner than most internal hosting of servers, with efficiency being a key focus of their business. Hence they qualify for investment in our Resource Efficiency theme. We are currently invested in Telecity – one of the leading European independent data centre operators, operating premium data centres with high levels of connectivity. These companies will continue to benefit from the rapid growth of data storage and are helping to make all this data less power hungry.

Recent posts

  • This year’s new killer
  • Seeing the bigger picture – Cooper Companies and myopia
  • What does 2020 hold for sustainable investing?
  • Politics playing catch-up on climate change
  • The great smog; London’s dirty air
  • Lessons from Woodford
  • Varian and the changing nature of cancer care
  • The bigger picture; declaring a climate emergency
  • Better out of it: the price of oil politics
  • Greta gets clean away
  • Archive

  • February 2020 (2)
  • January 2020 (1)
  • December 2019 (1)
  • November 2019 (2)
  • October 2019 (3)
  • September 2019 (1)
  • August 2019 (2)
  • July 2019 (3)
  • June 2019 (2)
  • May 2019 (3)
  • April 2019 (1)
  • March 2019 (1)
  • February 2019 (2)
  • January 2019 (3)
  • December 2018 (1)
  • November 2018 (2)
  • October 2018 (4)
  • September 2018 (2)
  • August 2018 (4)
  • July 2018 (1)
  • June 2018 (1)
  • May 2018 (1)
  • April 2018 (2)
  • March 2018 (2)
  • February 2018 (1)
  • January 2018 (1)
  • December 2017 (3)
  • November 2017 (1)
  • July 2017 (3)
  • June 2017 (1)
  • May 2017 (1)
  • April 2017 (1)
  • February 2017 (2)
  • November 2016 (1)
  • August 2016 (1)
  • July 2016 (1)
  • June 2016 (1)
  • May 2016 (1)
  • April 2016 (2)
  • February 2016 (1)
  • December 2015 (1)
  • November 2015 (3)
  • October 2015 (1)
  • September 2015 (1)
  • July 2015 (2)
  • April 2015 (2)
  • February 2015 (2)
  • December 2014 (2)
  • November 2014 (3)
  • October 2014 (4)
  • August 2014 (1)
  • July 2014 (3)
  • June 2014 (1)
  • April 2014 (2)
  • March 2014 (2)
  • February 2014 (3)
  • January 2014 (4)
  • December 2013 (4)
  • October 2013 (5)
  • September 2013 (3)
  • July 2013 (4)
  • June 2013 (2)
  • May 2013 (4)
  • April 2013 (2)
  • March 2013 (4)
  • February 2013 (6)
  • January 2013 (2)
  • December 2012 (3)
  • November 2012 (1)
  • October 2012 (4)
  • September 2012 (2)
  • August 2012 (1)
  • July 2012 (3)
  • June 2012 (3)
  • May 2012 (6)
  • April 2012 (4)
  • March 2012 (5)