Skip to main content
ScotlandComputer ScienceSyllabus dot point

What is the environmental cost of computer systems, and how can it be reduced?

The environmental impact of computer systems: their energy consumption, ways to reduce that impact, and the environmental considerations of intelligent systems.

An SQA Higher Computing Science answer on the environmental impact of computer systems, covering their energy consumption, practical ways to reduce that impact, and the environmental considerations raised by intelligent systems.

Generated by Claude Opus 4.810 min answer

Reviewed by: AI editorial process; not yet individually human-reviewed

Have a quick question? Jump to the Q&A page

Jump to a section
  1. What this key area is asking
  2. The energy cost of computer systems
  3. Reducing the impact
  4. Intelligent systems
  5. Examples in context
  6. Try this

What this key area is asking

The SQA wants you to understand the environmental impact of computer systems: their energy consumption, the practical ways that impact can be reduced, and the particular environmental considerations raised by intelligent systems such as large machine-learning models. This is a description-and-explanation topic, not a calculation one.

The energy cost of computer systems

So the environmental impact is dominated by energy use and the emissions behind it, but it is not the only factor. Manufacturing computers uses raw materials and energy, and disposing of old equipment creates electronic waste (e-waste), which can contain harmful substances. The full picture spans manufacture, use and disposal.

Reducing the impact

The impact can be reduced at every stage, and the SQA expects practical measures.

  • Energy-efficient hardware. Choose low-power processors and components, and equipment rated for efficiency.
  • Power management. Enable sleep or hibernate when idle, dim or switch off screens, and turn machines off out of hours rather than leaving them on.
  • Renewable energy. Power offices and data centres with renewable sources such as wind or solar, cutting the emissions per unit of electricity.
  • Better cooling. Use efficient or free-air cooling in data centres so less electricity is spent removing heat.
  • Virtualisation. Run several virtual servers on one physical machine so fewer physical servers are needed.
  • Recycling and reuse. Repair, reuse or recycle old equipment to reduce manufacturing demand and keep e-waste out of landfill.

Intelligent systems

Intelligent systems (artificial intelligence and machine learning) raise their own environmental considerations because they are exceptionally computationally demanding.

There is a balance, though. Intelligent systems can also reduce environmental impact elsewhere, for example by optimising electricity grids, routing transport more efficiently or improving building energy use. The SQA expects you to weigh the heavy energy cost of training and running such systems against the savings they may deliver.

Examples in context

These choices have real weight. Major cloud providers now report data-centre power use and increasingly buy renewable energy, and "green data centre" design (efficient cooling, high server utilisation) is a competitive selling point. E-waste is a growing global problem, which is why right-to-repair and recycling schemes matter. The energy cost of AI is now a live public debate: training the largest models can use as much electricity as many households use in a year, which is exactly the kind of environmental consideration this key area asks you to discuss.

Try this

Q1. State the main reason computer systems contribute to carbon emissions. [1 mark]

  • Cue. Their electricity use, much of it generated by burning fossil fuels.

Q2. State two practical measures to reduce the environmental impact of computers. [2 marks]

  • Cue. Any two of: energy-efficient hardware, power-saving settings, switching off when idle, renewable energy, efficient cooling, virtualisation, recycling.

Q3. Explain why training a large machine-learning model raises environmental concern. [1 mark]

  • Cue. It runs many powerful processors for a long time, using a great deal of electricity and producing significant emissions.

Exam-style practice questions

Practice questions written in the style of SQA exam questions on this dot point, with worked answer explainers. The year tag is the paper they imitate, not the source.

SQA Higher (style)4 marksDescribe two ways the energy consumption of computer systems contributes to environmental impact, and state two practical measures an organisation could take to reduce it.
Show worked answer →

Two contributions to environmental impact:

  1. Computers and especially large data centres draw large amounts of electricity, much of it generated from fossil fuels, producing carbon emissions.
  2. Data centres also require energy to cool the equipment, adding further to the electricity demand and the associated emissions.

Two practical measures:

  1. Use energy-efficient hardware and enable power-saving settings (sleep or hibernate when idle, switching machines off out of hours).
  2. Power data centres or offices with renewable energy, and improve cooling efficiency (for example free-air cooling) to cut the electricity used.

Other valid measures: virtualisation to use fewer physical servers, recycling old equipment to reduce manufacturing impact and e-waste. Markers reward two valid impacts linked to energy and two sensible reduction measures.

SQA Higher (style)3 marksExplain why intelligent systems, such as large machine-learning models, raise particular environmental considerations.
Show worked answer →

Intelligent systems such as large machine-learning models are computationally very demanding. Training a large model can require running many powerful processors for long periods, consuming a great deal of electricity and producing significant carbon emissions, far more than ordinary software.

They also rely on large data centres for both training and for answering requests once deployed, so the energy use (and the cooling it needs) continues in use, not just during training.

A consideration in their favour is that intelligent systems can also reduce environmental impact elsewhere, for example by optimising energy grids or transport, so the net effect must be weighed.

Markers reward the point that training and running such models is highly energy-intensive (large emissions), the reliance on data centres, and ideally the balancing point that they can also help reduce impact elsewhere.

Related dot points

Sources & how we know this