I am simulating flow through a 2D pipe that is getting smaller as it goes downstream.  Both simulations are using double precision, 2D solver, with Energy and standard k-e, standard wall fn and all the other models are off.  I have the same materials and boundary conditions in both simulations as well.  The outlet is pressure outlet with a target mass flow rate and preventing reverse flow, the inlet is also prevent reverse flow.  Both have no heat transfer properties on any of the walls.  All of the solution methods are the same and the monitors are set to the same values.  All of the values and options that I have put into the 19.3 solver on the schools virtual computer are identical to the inputs I have put in the 2020 solver on my computer.  

I have not attempted a simpler model yet to see if I get similar issues.  That will be the next step of trying to troubleshoot if I can't isolate another potential input that might be different between the two versions.  

Thank you for moving the question.