Examples
Worked examples
- Is an instance
A bioinformatics pipeline reporting its SCI score per sample processed, after caching a shared reference index instead of recomputing it on every run.
Counter-examples
Looks similar, but isn't
- Not an instance
Adding a large persistent cache that speeds up one metric while increasing total embodied and operational emissions from the extra storage hardware it requires.
Editorial commentary
Green software engineering applies eight principles including carbon awareness, energy efficiency, hardware efficiency, measurement, and climate commitments. In research it manifests as optimising hot inner loops, avoiding unnecessary precision, caching reusable artefacts, deleting unused training checkpoints, and selecting algorithms with lower computational complexity. The Software Carbon Intensity (SCI) specification published by the Green Software Foundation is the emerging measurement standard. Research software engineers (RSEs) are increasingly trained in these principles.
The Software Carbon Intensity (SCI) score
SCI expresses a rate — carbon emitted per one unit of a defined functional unit (per API call, per pipeline run, per analysis) — rather than a total, so that a system can be compared as it scales. Conceptually it combines the operational energy consumed multiplied by the carbon intensity of the electricity supplying it, plus a share of the embodied emissions of the hardware used, divided by the functional unit. Because it is a rate rather than a total, a change that increases total emissions while serving proportionally more requests can still improve (lower) the SCI score, and the reverse — this is a deliberate design choice in the specification, and worth stating explicitly when reporting it.
Worked example
Rewriting a genomics variant-calling pipeline step to use a lower computational-complexity algorithm and caching an intermediate reference index across runs, instead of recomputing it each time, reduces both wall-clock compute time and the SCI score for that pipeline’s functional unit (per sample processed) — a direct application of the “hardware efficiency” and “measurement” principles.
Counter-example
Adding an aggressive caching layer that speeds up a workflow but requires a much larger persistent storage footprint can increase embodied and operational emissions overall (extra disks, extra data-centre floor space) even though the visible compute time improves — a reminder that green software engineering weighs the whole system, not just the metric that is easiest to see.
Related pages
See also compute carbon footprint, carbon-aware computing, energy proportionality, sustainable HPC, and compute FLOPs estimation.
References
- Green Software Foundation, Software Carbon Intensity (SCI) Specification (greensoftware.foundation).
- Green Software Foundation, Principles of Green Software Engineering.
- Green Software Foundation, Carbon Aware SDK (github.com/Green-Software-Foundation/carbon-aware-sdk).
Also known as
Green software · Sustainable software engineering · Carbon-efficient software
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Green software engineering"
vocab-term-identifier="https://casrai.org/dictionary/term/green-software-engineering" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/green-software-engineering",
"name": "Green software engineering",
"identifier": "https://casrai.org/dictionary/term/green-software-engineering",
"description": "The discipline of designing, building, and operating software with explicit attention to its energy and carbon impact, as defined by the Green Software Foundation.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/sustainable-research#set",
"url": "https://casrai.org/dictionary/term/green-software-engineering",
"alternateName": [
"Green software",
"Sustainable software engineering",
"Carbon-efficient software"
],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"author": {
"@id": "https://casrai.org/#editorial-team"
},
"datePublished": "2026-05-21T02:23:02",
"dateModified": "2026-08-23T05:19:53",
"inLanguage": "en-GB",
"isAccessibleForFree": true
}
