The problem

Efficiency makes computing cheaper, and cheaper computing invites more use. Without a limit, the saving is spent on more load and emissions don't fall: the rebound effect.

The fix you can use today

• Pair efficiency work with an energy or carbon budget.
• Retire the capacity an optimisation frees up.
• Report absolute energy, not only efficiency per unit.

Researchers led by Lancaster University, with Mike Berners-Lee, argue that savings only count when total use is capped. GreenCode's climate case builds on this work.

Tools that help

• Software Carbon Intensity (SCI) (Green Software Foundation): the Green Software Foundation's method, SCI = ((E x I) + M) per R, adopted as ISO/IEC 21031:2024
• Cloud Carbon Footprint: estimates the energy and emissions of AWS, Azure and Google Cloud usage from billing and usage data

How GreenCode helps

GreenCode helps tackle this by measuring reductions in absolute terms, not only per task. Its climate case shows owners where capacity can be retired, so a saving can be banked rather than spent. More on this.

Tags

#governance#esg#cost

Sources

• The climate case: rebound effects and net benefit (Widdicks, Berners-Lee et al., Patterns 2023)
• The Shift Project, Lean ICT: our new report: 'buy the least powerful equipment possible, change them as rarely as possible, and reduce unnecessary energy-intensive uses'
• W3C, Web Sustainability Guidelines: environmental budget across the lifecycle

All GreenCode Tech Tips