Nvfortran vs pgfortran?

Forgive me if this is the wrong forum for this. I’m a little confused as to what the difference is between the two compilers. I’ve been writing CUDA FORTRAN code and so far I’ve just been testing it on a cluster which has pgfortran installed. Unfortunately not every machine has pgfortran installed whereas far more people I know how nvfortran installed, so I was wondering what the difference is and what I need to be careful of if I’m using one but not the other.

Can code written and compiled on one be compiled on the other? Or are they for different things and work completely differently? I haven’t really been able to find any clear answers in all the documentation I’ve read.

They’re really the same compiler just a different name.

PGI was rebranded to be the NVIDIA HPC Compiler when we decided to pull together multiple NVIDIA products into the NVIDIA HPC SDK. Though we kept the PGI compiler drivers as part of the package so we didn’t break makefiles as we made the transition.

So you can use either, but the official driver going forward will be nvfortran.

I see, that simplifies affairs quite a bit. Thanks again for your help Mat, at this point I’m starting to think I should just take my name off of this CUDA code and put yours on it for all the help you’ve given me these past few months. Well, your and Robert’s anyway.