Need to run the golden reference model with same set of inputs that are fed to the DUT and compare DUT monitor results with reference model results.
In reply to Abhishek S:
You should be able to duplicate the agents connected to your reference model and DUT. Then connect the monitors analysis ports to the expected and actual analysis exports respectively.
If you have separate input and output agents, you might not have to duplicate the input agents, and just tie the inputs together.
If your two models are not 100% cycle, accurate with respect to each other then you’ll have to duplicate the sequences sent to each agent.
In reply to dave_59:
Considering the first suggestion, to duplicate the agents. If the data inputs are of rand type, how do we ensure that randomization shall generate same values from both agents ?
In reply to Abhishek_S_Reddy:
You put one of the agents in passive mode and make continuous assignments from the active interface to the passive interface.
It is hard for me to provide more advice without more details of your reference model compared to your DUT.