Speeding up compilation of template-heavy code

[Per Nordlöw via Digitalmars-d-learn]

We're looking for a way to speed up compilation of template-heavy 
code. So we are trying to find out which parts of the code that 
is most costly to compile.

Is there a dmd flag that shows the code after template 
instantiations has been performed? Or some other dmd flag that 
can help out finding hot-spots in the compiler in our case.