Green Software Engineering Principles — Vocabulary

Green Software Foundation (greensoftware.foundation) three pillars: Carbon Efficiency (emit the minimum amount of carbon to deliver your function — demand and spatial shifting, renewable energy), Energy Efficiency (use the minimum energy — algorithmic efficiency, resource utilization), Hardware Efficiency (extend hardware lifespan, avoid unnecessary hardware creation — embodied carbon).

SCI = (E x I) + M per R. E = energy consumed, I = carbon intensity of energy, M = embodied carbon of hardware (amortized), R = functional unit (per API call, per user, per transaction). SCI allows comparing carbon efficiency: 'Our API has an SCI of 5gCO2eq per 1000 requests. After optimization, it dropped to 3gCO2eq per 1000 requests.'

Right-sizing: if your application uses 10% of a large instance, switch to a smaller one. Wasted compute capacity = wasted energy = unnecessary carbon emissions. Cloud providers offer tools (AWS Compute Optimizer, Azure Advisor) to identify over-provisioned resources. Right-sizing is usually the highest-impact low-effort green action.

Embodied carbon: manufacturing a laptop or server requires significant energy (often from fossil fuels), producing CO2 before the device is ever turned on. For a laptop, embodied carbon may represent 80% of its lifetime carbon footprint. Extending hardware lifespan, buying refurbished equipment, and avoiding unnecessary hardware procurement reduces embodied carbon significantly.

Carbon budget for software: 'This new ML feature has a carbon budget of X gCO2eq per day. If the initial design exceeds this, we must optimize before shipping.' Treating carbon as a constrained resource — like memory or cost — makes sustainability a first-class design constraint rather than an afterthought.