We can categorize the tools & APIs into two dimensions:
- that help us assess the embodied impact of the equipment and infrastructure (static)
- that help us determine the operational impact stemming mainly from energy & water usage (dynamic)
Embodied Impact (Static)
Embodied impacts of a physical server type
Get the impacts of a server - Boavizta API documentation
https://doc.api.boavizta.org/getting_started/single_server/
Embodied impacts for an AWS instance type
Get the impacts of a cloud instance (AWS) - Boavizta API documentation
https://doc.api.boavizta.org/getting_started/single_cloud_instance/
Operational Impact (Dynamic)
Measuring on physical machines
Note that these tools will not work in Cloud or Virtualized environments as they require root-level access to the machine to call the RAPL or IPMI interfaces.
scaphandre
hubblo-org • Updated Feb 14, 2023
Measuring using different methods (both virtualized and physical)
The green metrics tool supports different ways to determine energy usage, including the SDIA Formulas and a machine learning model based on Spec SERT.
green-metrics-tool
green-coding-berlin • Updated Feb 14, 2023
Estimating power usage of a CPU via machine learning model
spec-power-model
green-coding-berlin • Updated Jan 13, 2023
Reference Testing
If you need to test a piece of code or a repository on a reference system in which all measurements are available, you can create Gitlab repository in our laboratory and run it on our machines.