It's come up that some simulators ignore the initial delay in the rising/falling waveforms of an IBIS model - i.e. the flat portion of the waveform before any change in voltage happens. Does HyperLynx use the initial delay correctly or does it ignore it?
This is relevant for using HL to compare delay across two different paths, using two different IO standards that otherwise model Tco correctly as the IBIS rise/fall waveform initial delay. (the application here is DDRx memory, DQ vs DQS or A/C vs CK)