/myDict<<
           /myProc {
               (Hello world) =
           }
>>def

myDict begin %A
myProc       %A
end          %A

myDict /myProc get exec %B

Given the above Postscript code, which (not surprisingly) produces the following output using ghostscript (GPL Ghostscript 9.55.0 (2021-09-27) on Linux):

Hello world
Hello world 

The %A lines pushes the whole dictionary and any state it might have onto the dict stack; I don’t think this is true for the line marked %B, as I believe it just does a lookup and runs the procedure it finds, without loading the dictionary onto the dict stack .

I am interested for times where the dict is very large and loading it (perhaps many times) would have significant overhead.
Context: Running in ghostscript.

Question: Does the last line (myDict /myProc get exec %B) have less overhead than the lines which are marked with %A?