Is there any language in which a type named double occupies less space than the type name float?
I doubt it. The name double refers to the precision. It means that it has twice the precision as a float. Storing more precision requires more space.
double