I am not sure why there should be differences.
It would be good to check the behavior of the material models by doing, say one or two cycles on a tension test using a single element. I do not recall making changes in the model you are using. In finite strain you cannot use kinematic hardening with the isotropic elasto-plastic model (it is a blunder in the original development).
Next check that the two programs give the same final equation set. You can do this by using the command SHOW ID and look at the equation numbering. It could be that the linking is different in the two versions. But a careful check that you get the same problem before you run the multicycles is needed.
The input file seems to be running correctly for me -- I have not checked the actual load history, but since you are summing reactions to get the force any discontinuity should be evident to you.
