I suspect this is the parser taking its time processing the 2GB of data. Can you see any threads taking up a lot of CPU?
Also, how many atoms are there in your system? IQmol relies on OpenBabel to identify rings and bond types in the system, but currently this step does not scale well with system size and so if you have a large number of atoms in your system it can take a long time.
If you post the contents of the iqmol.log file produced after trying to load the file, I might be able to see if there is any other problem.
I’ve had some 0.5-1.0 Gb files with NTOs from a large TDDFT calculation, it’s been a while but I recall that those take something like 10-20 min to load in IQmol, at first I thought it wasn’t working until I went for a coffee.