IEEE comments on DC sustainability

With Gartner predicting that 75% of organizations will need to have implemented a data centre infrastructure sustainability programme by 2027, Aoife Foley, IEEE (www.ieee.org) Senior Member and Profession in the School of Mechanical Aerospace Engineering at Queen’s University Belfast believes that poor storage practices or ‘dark data’ will be a massive drain on resources in a market that is already disrupted by price fluctuations and supply constraints, and that organizations can achieve greater resilience and better risk management by optimizing server utilization and storage capacity.

“Dark data accounts for about 54% of data that’s stored around the world, most of which has no function – all that data comes at a cost.  When you look at the way the data is all gathered, it’s unstructured.  Storing a massive amount of dark data wastes energy, most of which is powered by non-renewable resources.  For example, in the United States, power consumption due to data centre data storage was estimated to be at 14 billion kWh in 2020 resulting in almost 6.5 metric tons of CO2 emissions.”

“As Gartner’s latest report outlines, achieving sustainability means addressing environmental considerations during solution design as well as during the build.  Solutions must meet pre-defined and agreed environmental sustainability criteria.  Sustainable data centres require designs that consider energy conservation, energy efficiency and renewable energy sources.  Businesses and data centres should start by simply taking control of data storage, assessing the storage rules, and ensuring they are not holding data that is no longer needed.”

“It is essential that leaders improve their data management policies, identifying which data is in fact valuable, and eliminating any dark or redundant data from their data centres to avoid emissions spiralling out of control and avoid unnecessary digital waste.  Leaders should also enquire about their data centre provider’s multi-site footprint and its ability to enable distributed network availability and dynamic load placement.  Identifying workloads that can be transferred from peak demand periods to off-peak hours ultimately results in lower costs for all parties.”