Fortran was still needed for physics in 1990. In fact, it is still needed in many subfields of physics in 2017.
You are right that it is used still. Happily/sadly (depending on your perspective) there's a "new" version of Fortran coming out soon called Fortran 2015 (coming out mid-2018 because it takes this long to write the first program since its release).
I never understood why c++ is not used for serious physics calculations given how much easier it is to write stuff, but there are people who are very sure that Fortran is superior to anything else. I just wish that Fortran would have a good way to print out results so you can see what is being calculated properly!