- [ ] Add a visual of the entire Cloud stack from a server-side application
To optimize energy efficiency:
- Total power consumption (including the overhead of cooling, etc.) of the container, physical or virtual server on/in which the software runs
- Quantitive vs. qualitative
- Fixed constants: power consumption at idling levels (120 kW at 10%, 140 kW at 20%, 150 kW at 30%, ...)
- % idling capacity (memory, compute, and storage) of the underlying hardware
To address its carbon emissions:
- Current carbon intensity of the power supplied to the underlying data center
- is the backup generator running? ⇒ change carbon intensity
- is feeding on the grid? ⇒ change carbon intensity
- coverage by certificates / PPAs
- Current pollution of the power supplied to the underlying data center (if any)
- Embedded emissions in IT infrastructure (servers, switches) and underlying data center (cooling, electrical, building) and network infrastructure
- Active power consumption of all attached network equipment
Enrichment Information
Future data points
Forecast & load curve
- Renewable power forecast (local to facility/grid zone)
Constraints from the infrastructure
- Ludger: Security of the networks (infra vs. IT networks)
- JohnB: EU CoC, security of the data, the data is there
- BCS has released a fixed measurement for IT load
- Lobby GHG scope protocol
- ICT guidance
- SBTI has replaced IT