- ✓Data centres consume approximately 200 terawatt-hours of electricity globally each year, making energy efficiency a critical concern: best practices include server virtualisation, hot/cold aisle containment, free cooling and power usage effectiveness (PUE) optimisation.
- ✓Hardware procurement decisions have significant sustainability implications: devices with longer support lifecycles, higher repairability scores and responsible supply chain practices represent a substantially lower lifetime environmental impact than cheaply made, short-lived alternatives.
- ✓The circular economy principles of reduce, reuse and recycle apply directly to hardware management: extending device lifecycles, refurbishing and redeploying equipment, and ensuring responsible recycling of end-of-life devices are all practical sustainability measures.
- ✓Software efficiency has a direct impact on energy consumption: applications that are poorly written or unnecessarily resource-intensive consume more CPU, memory and network bandwidth, requiring more physical infrastructure and more energy to run.
- ✓Renewable energy procurement for digital infrastructure, either through direct power purchase agreements with renewable generators or through the purchase of renewable energy certificates, is an increasingly common and important element of corporate digital sustainability strategies.
Listen to the full episode inside the course. Enrol to access all 80 episodes, plus assignments, tutor support and Student Finance funding.
Start learning →Alex: Hello and welcome back to The Study Podcast. Today we're looking at Green IT in practice: what organisations actually do to reduce the environmental impact of their digital operations. Sam, this is where principles become action.
Sam: And where it becomes clear that this isn't just about carbon offsets or renewable energy certificates: there are real, concrete engineering and procurement decisions that organisations make that determine their actual environmental footprint.
Alex: Let's start with data centres, since they're typically the single largest source of energy consumption in digital operations.
Sam: Data centres face pressure on multiple dimensions: the energy they consume to run servers, the energy needed to cool those servers, and the source of that energy. Server virtualisation, which allows multiple workloads to run on the same physical hardware rather than dedicating a server to each application, dramatically improves utilisation and reduces the total hardware needed. Containerisation takes this further. Hot aisle and cold aisle containment is a data centre design practice that keeps hot exhaust air separate from cold intake air, reducing cooling requirements significantly. And free cooling, using outside air or water when ambient temperatures allow, can dramatically reduce the energy needed for cooling in appropriate climates.
Alex: And the metric PUE, power usage effectiveness, is how data centre efficiency is measured?
Sam: Yes. PUE is the ratio of total data centre energy consumption to the energy used by the IT equipment itself. A perfect score is 1.0, meaning no energy is used on anything other than computing. A PUE of 1.2 means twenty per cent of energy goes to cooling and other overhead. Google and Microsoft now operate some hyperscale data centres with PUEs below 1.1, which is remarkable. Older enterprise data centres often have PUEs of 1.5 to 2.0, representing significant inefficiency.
Alex: What about hardware procurement and lifecycle?
Sam: The environmental impact of electronic devices is heavily front-loaded: a large proportion of the lifetime carbon footprint of a device comes from its manufacture rather than its operation. This means that extending device lifetimes, by choosing hardware with longer manufacturer support, by repairing rather than replacing, by refurbishing and redeploying equipment rather than discarding it, is one of the most effective sustainability measures available. The circular economy principles of reduce, reuse and recycle apply directly.
Alex: And software efficiency is less talked about but matters too?
Sam: Significantly. Inefficient code consumes more CPU, memory and network bandwidth, which translates directly into more physical infrastructure and more energy. The concept of green software engineering is emerging as a discipline: writing software that accomplishes its purpose with the minimum necessary computational resources. This isn't just about code optimisation: architectural choices like choosing an appropriate database for a use case rather than a generic one, or running batch jobs during times when grid electricity is cleanest, also contribute. The Green Software Foundation has published principles and frameworks to guide this work.
Alex: And renewable energy procurement?
Sam: Major cloud providers and large technology companies have made significant commitments to matching their energy consumption with renewable sources, either through direct power purchase agreements with renewable generators or through renewable energy certificates. For smaller organisations, choosing cloud providers with strong renewable energy commitments is one of the highest-impact sustainability choices they can make.
Alex: Really practical. Thanks, Sam. We'll look at building a digital sustainability strategy next.