How much energy does the training needs?

 Training our model is more efficient using the old RTX 2070 then with the new RTX 4090.

RTX 2070 uses 166W of a maximum allowed of 175W. And the RTX 4090 uses  405W of the 450 maximum allowed during half the time (for the same information quantity).

It gives 21% more efficiency.





Comments

Popular posts from this blog

Celebrating Two Years of ChatGPT: A Virtual Companion Changing our World

Who was Ada Lovelace?

Can you tell us a story, Stella?