I was about to comment how this number doesn't make sense, but reading the article they mention the power savings for phones specifically is actually 2.2TWh per year, which might be realistic. The rest of the 14 TWh comes from landline phones (2.2TWh) and energy used to produce phones that would be saved (8.1 TWh) from the lifetime extension measures, that aren't even related to the power efficiency of phones and won't take place in the EU.

However, even these numbers are overinflated when you take into account they're using a PEF factor of 1.9, meaning they multiply the actual power usage by 1.9 to adjust for stuff like power transfer losses (only 5-10%) and the inefficiency of generating power from e.g. fossil fuels (because e.g. petroleum might only convert 40% of its potential energy to electricity), but when people typically talk about power usage they're talking about the actual amount of electricity that needs to be generated, not this abstract representation of it, meaning that e.g. phone electricity savings are actually only 1.1TWh-1.2TWh.

I was thinking at first that this would lead to even more aggressive "app optimization" where it would refuse to notify alarms you've explicitly set due to battery life concerns.

But instead there's a better way, which I hope manufacturers will take. Perhaps smartphone chips can stop chasing absolute peak performance and instead focus on good performance at a reasonable power budget.

Of course the biggest problem is bloated software but idk how we can fix that.

I think the big value of these labels are not the energy class (like for bigger appliances) it’s the fields under the energy class. Repeatability, (official declaration of) battery duration, simple to spot resistance to dust/water, etc

1.6MW is absolutely tiny for a nuclear reactor and most produce at least 200MW with a lot of them going up to a GW and not to say 14TWh isn’t a huge amount of energy just that the comparison to a nuclear power plant is off

... optimise the use and recycling of critical raw materials

Does anyone have an idea how this helps with recycling?

I’m having trouble finding actual ratings for anything on this site

Would be funny to know how the supposed 14TWh savings compare to the cost required to introduce/maintain these labels. And by costs I mean overall costs with everything related to it.