Using the correlation workflow, you can compare the results of the
IBIS-AMI models generated using SerDes Toolbox with the original Simulink models.
Open the SerDes Designer app. Add a FFE to the transmitter side and a CTLE
and DFECDR to the receiver side.

Generate the dll/so and ami models by clicking the Export >
Make IBIS-AMI model for SerDes System in the app toolstrip. Export
the SerDes system to Simulink.
In the Simulink model, add a Delay between the Rx subsystem and the
Eye Diagram block. Insert an AMI block and another
Eye Diagram and connect them in parallel with the branch containing the
Rx subsystem, Delay, and Eye Diagram blocks.

Double click the AMI block to open the Block Parameters dialog box.
Refer to the dll and ami files for the receiver you created before. The block
automatically creates the Input String parameter and populates the
AMI tree.

Change the Operation parameter to Rx
and select the Correlation Flow option.
Since you are correlating Rx subsystem to the previously generated IBIS-AMI model
using the AMI block, double check that the settings for the two systems
match.
The AMI block introduces a 1024 sample delay for compatibility with the
block size requirement of the IBIS-AMI standard. If you are not doing sample by sample
correlation, you can ignore this delay.
Since this example uses sample by sample correlation, you need to mitigate the delay
introduced by the AMI block. To do so, set the Delay
length of the Delay block to 1024.
Enable data logging for the outputs of the AMI and Delay
blocks by selecting the signals and clicking the Log Signals button
in the Simulation tab.
Run the simulation. Open the Simulation Data Inspector and select the AMI and Delay
signals. The two signals overlay perfectly.

To compare the two signals, click the Compare button. Select
both signals and set Global Tolerance parameter to
1e-12 to use a small voltage tolerance of 1 pV.

Observe in the bottom panel that the difference is well within the tolerance set. Both
the Rx subsystem and AMI block outputs correlate extremely well.