OPL fails at building very large scale models due to the fact that it uses a "row api" of concert : Variables are created and added then to the range constraints.
Unfortunately, such API can become dramatically slow for very large models, and building a model that runs sometimes in a few seconds can take a very long time (sometimes up to several hours !). Why is the concert generated code that fills cplex model not using the column api instead ? This one is known to be much faster and we can see a huge tradeoffs for such large models
Note that this choice is totally transparent for the OPL developer, so there is here no impact of such change for OPL users
NOTICE TO EU RESIDENTS: per EU Data Protection Policy, if you wish to remove your personal information from the IBM ideas portal, please login to the ideas portal using your previously registered information then change your email to "email@example.com" and first name to "anonymous" and last name to "anonymous". This will ensure that IBM will not send any emails to you about all idea submissions