Skip to content

CO2 estimation on the clusters#

Current computation#

On SCITAS clusters you are presented with an estimated quantity of CO2eq of your computations

The CO2eq estimation is currently only based on power consumption of the machines. It is not an exact measure of the job consumption but it is based on an average power draw of the machines.

  • We get the power draw per rack in the different Data centers (CCT and INJ). This power draw include the compute nodes, but also the network, scratch storage, front nodes and admin nodes for each cluster.

  • This power draw is multiplied by the Power Usage Effectivness (PUE), also provided by the data center team, in order to account for the cooling of the machines.

  • Finally the energy consumption is converted in g CO2 eq by using a conversion coefficent based on the energy mix from Switzerland.

From all this measure we derive a single coefficient per cluster of g CO<sub>2</sub> eq / core hour or g CO<sub>2</sub> eq / gpu hour

This data and coefficent can be found in the following Google Spreadsheet

The currently used factors are

Cluster W/(core or GPU) g CO2 eq / (core h or GPU h)
helvetios 8.19 1.11
izar 308.8 41.88
jed 6.93 1.02

Futures possible improvements#

  • In the average coefficients we should account for the CO2 eq of the construction of the machines.

  • We are also working on measuring the power consumption of the job with hardware counters, this will account for CPU, GPU but will most probably use heuristics for RAM, storage and network.

  • We also want to evaluate the impact of data storage on the shared storage /home, /work, /archive, ...