Green Data Strategies to Cut Corporate Energy Costs and..
Date: 07 JUNE 2022
Author: Bulk Infrastructure AS
Over the past few years, the financial and fintech sector has seen rapid growth in both AI and machine learning. Unfortunately, this leads the industry into difficulties like the vast energy consumption required to process this data.
Plus it creates the need for paring the right IT technology with a scalable, eco-friendly, energy-efficient data center infrastructure which is needed to store and process such huge volumes of data.
Addressing these issues is critical when it comes to raising future investment and meeting environmental, ESG, SDG and sustainability targets.
Warren showed the global data consumption forecast until 2025, demonstrating a threefold increase in data consumption from 64 zettabytes in 2020 to more than 180 zettabytes by 2025.
The knock-on effect is that more servers will be needed to store larger data sets and thus more energy will be needed to keep them running – resulting in a far greater output of carbon emissions… all of which is catastrophic for the planet.
The Growth of CO2 Emissions is driven by the increased use of Artificial Intelligence
The graph below compares the CO2 emission footprint of taking a flight from NYC to San Francisco, with an average human life (in the US and globally), with the manufacturing of a US car and finally the training of an AI model. This shows visually how much energy is needed in modern technology environments where AI models are being trained.
Shockingly, the carbon footprint of training just one AI model is more than that of 56 people in their whole lifetime combined.
In addition, global electricity consumption is forecasted to double, and greenhouse gas emissions will triple by 2030.
The Carbon Footprint of Data Centers can’t be ignored
As a result, data centers could end up accounting for 13% of annual global electricity consumption and 6% of the carbon footprint. Therefore, responsible businesses place these workloads in markets that deliver 100% renewable energy.
Regions that can guarantee 100% zero carbon emissions help to reverse these negative greenhouse gas emission trends.
One of the best solutions for reducing this impact is to move data workloads to a region such as Norway, because it runs entirely on 100% renewable energy.
Every 1 MW workload in a London data center results in the same emissions as 12.8 MW in Norway, a vast increase. Metaphorically this is the equivalent “Carbon output” of 2,137 cars being driven for a year for each 1 MW of energy consumed.
It is also important to remember that not all workloads need to be in urban areas. The graphic below compares the carbon impact between London and Oslo in more detail.
Why Geographic Decisions are key to sustainable Green Data Strategies at Low Cost
Warren further compared the electricity consumption and carbon emissions through ElectricityMap.org which provides worldwide access to 24/7 grid carbon intensity historically, in real time and as forecast for the next 24 hours.
Through this website, Norway can be compared to other key European countries. It was seen that the Southwest of Norway produces 21 grams of carbon for every kilowatt of energy produced whereas other European markets like the UK produce 287 grams of carbon.
Germany produces 292 grams, France produces 81 grams and Spain produces 84 grams of carbon for every kilowatt. So what does this mean?
These statistics paint an interesting picture and argument as to why moving data workloads to regions like Norway where power is 100% carbon free can not only lower operations costs by 60-80% but also reduce a business’ carbon footprint.
In fact, according to the case study about the energy cost difference between running the same infrastructure in London versus running it in Norway, the energy prices in Norway are almost 80% less than energy costs in London, illustrated by the chart below:
Note: The data presented was taken in March 2022.
The case study showed 78% lower costs which represented a saving of 225 million Euros over a 10-year term.
In summary, the most successful green data strategies respond to the huge increase of energy required by AI intensive industry sectors such as financial services and fintech by incorporating decisions on the geographic distribution of data centers.
The “use case” for moving data centers to green and low energy regions like the Nordics certainly makes sound environmental and economic sense.
Watch the Full Webinar here >>
For more information on Bulk’s data center offering for financial services and fintech companies, please visit Bulk’s website here.
