Validate the energy / carbon calculation methodology

Hi Maxime,

Thanks for the initiative!

I'm a bit surprised about the example screenshot in your README: 4.5 MWh for a single machine is just enormous. Assuming it runs continuously at 180 W, this would be the energy consumed in almost 3 years. Did you validate your calculation method in simple cases to check that it gives you the right orders of magnitude?

Also, if you only use CPU usage, you miss a big part of the carbon footprint (manufacturing phase of the server, cooling, the need for infrastructure servers to manage the platform e.g. storage servers...)

This publication goes into much more details: https://hal.science/hal-04762718