Energy efficiency leading to more simulations?


I need your brain :slight_smile:

Here’s cool repo where energy efficiency is estimated for different languages.

I remember discussing about an effect “that-has-a-name-I-can’t-remember” where it states that using a more efficient code/language/etc does not lead to less energy consumption because people just do more simulations instead of keeping the number of simulations constant.

Does anyone knows the name of this effect? I remember reading something about that on Wikipedia, I just can’t find the right google search to re-land on this concept! There is the concept of “Induced demand”, but from memory, it was something related to computing. I might be wrong though!



And the very related Downs–Thomson paradox - Wikipedia


Thanks! I knew the first letter was “J” ! :smile: