Cloud Sustainability Vocabulary | Cloud FinOps Exercises

1 / 5

An architecture review included a cloud sustainability analysis.

What does cloud sustainability analysis typically cover?

Cloud sustainability analysis quantifies the carbon impact of your workload and identifies technical actions to reduce it.

Cloud provider tool	What it measures
AWS Customer Carbon Footprint Tool	Estimated CO2e from AWS services; trend over time
Azure Emissions Impact Dashboard	Carbon emissions per Azure region and service
GCP Carbon Footprint (console)	Carbon data per project and region; CFE% per region

Scope 2 = electricity consumed by cloud infrastructure. Scope 3 = supply chain emissions (hardware manufacturing). Key vocabulary: "cloud carbon footprint", "CO2e (CO2 equivalent)", "Scope 2 emissions", "sustainability analysis", "carbon-aware workload".

2 / 5

The team is selecting a cloud region with the lowest carbon intensity for a new flexible workload.

What is carbon intensity in a cloud context?

Carbon intensity measures how "clean" the electricity grid is in a given location — it varies by time of day and season, not just by region.

Region type	Carbon intensity	Primary source
EU-WEST-1 (Ireland)	Low	Wind energy
US-WEST-2 (Oregon)	Low	Hydro power
Coal-heavy grids	High	Fossil fuels

Google publishes Carbon-Free Energy (CFE) percentage per region. Key vocabulary: "carbon intensity", "grid emissions factor", "renewable energy mix", "low-carbon region selection", "carbon-free energy (CFE)".

3 / 5

The engineering team is adopting carbon-aware computing for their ML training pipeline.

What does carbon-aware computing mean?

Carbon-aware computing exploits the fact that grid carbon intensity varies by time (time-shifting) and location (location-shifting).

Strategy	Technique	Suitable for
Time-shifting	Schedule jobs when grid is cleanest (often nights/weekends in renewable-heavy regions)	Batch ML training, ETL, backups
Location-shifting	Route flexible workloads to the region with the lowest current CFE	Global distributed workloads

Tools: Electricity Maps API, Google carbon-aware scheduler, Carbon Aware SDK (Green Software Foundation). Key vocabulary: "carbon-aware scheduling", "time-shifting workload", "grid carbon intensity hourly", "Green Software Foundation", "Carbon Aware SDK".

4 / 5

The sustainability report includes a cloud efficiency metric for each workload.

What is cloud efficiency in the sustainability context?

Cloud efficiency in sustainability = doing more with less energy. Two levels: data centre efficiency (PUE) and workload efficiency (code/architecture).

Metric	Definition	Benchmark
PUE (Power Usage Effectiveness)	Total DC power ÷ IT equipment power	Hyperscaler: 1.1–1.2; enterprise DC: 1.5+
Workload carbon intensity	gCO2e per unit of business value	Lower = more efficient

Code optimisation reduces compute → reduces energy → reduces carbon. Hyperscalers' PUE advantage means cloud is often more efficient than enterprise data centres. Key vocabulary: "Power Usage Effectiveness (PUE)", "workload carbon intensity", "compute efficiency", "emissions per useful work".

5 / 5

The team committed to reducing cloud carbon emissions by 30% without reducing business capability.

Which technical strategies can achieve this?

Technical carbon reduction actually reduces emissions. Carbon offsets merely compensate — they do not reduce the energy consumed.

Technique	Carbon reduction mechanism
Rightsizing	Less compute running = less electricity consumed
AWS Graviton (ARM)	Up to 60% better performance-per-watt vs. x86
Storage lifecycle (S3 → Glacier)	Cold storage uses less energy per GB than hot storage
Low-carbon region selection	Same compute, lower grid emissions factor
Dev/test idle shutdown	Stop environments after hours; auto-start for CI runs

Key vocabulary: "Graviton efficiency", "ARM-based compute", "storage lifecycle carbon reduction", "technical carbon reduction", "embodied carbon vs. operational carbon".