Thank you for your question. Can you clarify what specific 'GMM model' you are referring to? Are you referring to the generalized method of moments(GMM) which is supported in SAS/ETS procedures PROC MODEL and PROC PANEL, or are you referring to the Gaussian mixture model(GMM) as supported in the Data Mining and Machine Learning procedures PROC GMM as documented here:
https://go.documentation.sas.com/?cdcId=pgmsascdc&cdcVersion=9.4_3.5&docsetId=casml&docsetTarget=cas...
These two models are completely different type of models although both have abbreviated name of 'GMM'. Please clarify exactly which GMM model you are interested in.
The FMM procedure is a SAS/STAT procedure designed for 'finite mixture models'(FMM) where the distribution of the response is a finite mixture of distributions. It is not the same type of models as the generalized method moments as supported in PROC MODEL or PROC PANEL.