How green is my data? – Part I

How green is my data? – Part I

Designed to maximize energy efficiency and minimise environmental impact, green data centres can proactively monitor our journey towards a sustainable tech future

Data centres are an essential component of our technological future. They are the gatekeepers of nearly all of the global Internet traffic. However, like most good things, data centres come at a price – power consumption, and the resulting greenhouse gas emissions, being a global concern. According to estimates, digital technologies were responsible for 4% of greenhouse gases emitted in 2019.About 1% of the world’s electricity demand, or 200–250 Terawatt hours, was consumed by data centres in 2020. And, by 2030, AI alone could consume around 20% of the world’s total electricity produced.

All this doomsday information means data centers,along with any associated activities, need to proactively monitor environmental impacts to sustain themselves. With the exponential growth and usage of the Internet, power consumption in data centres has increased significantly. Due to the resulting environmental impact, increase in public awareness, higher cost of energy and legislative action, there is increased pressure on companies to follow a green policy. And the most acceptable solution could be green data centres with efficient power and water utilization. A green data centre is a repository for storage, management and dissemination of data in a way where mechanical, lighting, electrical and computer systems are designed to maximize energy efficiency and minimise environmental impact.

Different metrics, one aim

How efficiently power is used in data centres is assessed through Power Usage Effectiveness or PUE, developed in 1997. This is calculated as the total power used by a data centre divided by the power consumed for computing tasks. A result of 1 means all of the power used by the data centre has been utilised for computational purposes – and none has been lost in transmission, or used for cooling or lighting purposes. This indicate a highly energy-efficient, and hence sustainable, data centre.

Image: Power Usage Effectiveness metrics;

Carbon Usage Effectiveness or CUE is another metric that wasdeveloped by The Green Grid, which is part of the Information Technology Industry Council. CUE is the ratio of carbon dioxide emissions a data centre generates divided by the energy consumption of the data centre’s equipment. The aim is to maintain the lowest possible value.

Yet another approach to evaluate the efficiency of data centres is based on Water Usage Effectiveness or WUEmetric. This is estimated as the annual usage of water used for humidification and cooling purposes in a data centre divided by the total annual kilowatt-hours (kWh) used to power IT equipment in that data centre. A lower WUE value denotes higher water efficiency.

Why water – you may ask! Although it may appearunrelated, water consumption is very much an issue in running data centres. All machine computation generates heat. Water is a scarce resource, and for cooling purposes alone a lot of it is required. A recent joint research by four US scholars has revealed that heavy computation-intensive activities like AI algorithms are veritable water-guzzlers. Going by WUE estimates, an AI model like OpenAI’s ChatGPT would consume 500ml of water to conduct a simple conversation comprising anything between 20 and 50 questions. Now reflect on how many users are currently using the very popular ChatGPT around the world at any point in time – and how many more would be using it once it becomes as mainstream as Google searchin near future–and you will realise why water consumption by data centres is going to be a huge environmental threat.

Green strategies to be adopted

The usual strategies a data centre can adopt to stay green are as follows:

  • minimized building footprint
  • low-emission building materials, carpets and paints
  • sustainable landscaping
  • e-waste recycling
  • catalytic converters on back-up generators
  • alternative energy, like photovoltaic technology, heat pumps and evaporative cooling technology
  • use of hybrid or electric vehicles.

Plan at the design phase

Whether upgrading an existing data centre or building a new one, energy efficiency and environmental considerations are essential design components. Ideally, a green data centre must focus on environmental aspects right at the design stage. It is much easier as well as cost-effective to incorporate elements that take care of energy efficiency, scalability, and ecological impacts of a facility during its construction phase. Organisations planning a green data centre can engage design consultants experienced in designing energy-efficient and environmentally friendly buildings.

[To be concluded]

Know more about our Top Ranked PGDM in Management, among the Best Management Diploma in Kolkata and West Bengal, with Digital-Ready PGDM with Super-specialization in Business AnalyticsPGDM with Super-specialization in Banking and Finance, and PGDM with Super-specialization in Marketing.

PGDM in business analytics



© 2023 Praxis. All rights reserved. | Privacy Policy
   Contact Us
Praxis Tech School
PG Program in Data Science