sarmas@TOPAZ.UUCP (E. Sarmas,,,) (03/15/86)
My experience has shown that esim is a very wierd tool to use if one wants to simulate large circuits. The problem is that esim does not iterate the simulation until a new stable state emerges. For example if you have an inverter whose input is IN and output is OUT and OUT drives other circuits and say you change IN from 1 to 0 then OUT will go from 0 to 1 but the OUT line that controls the other circuits will stay at 0 in esim's mind. This behavior may be right for those who wrote esim because it eases the simulation process (and the programming) but it is unrealistic and really breaks down the simulation when Xs start to appear due to classes betwwen 0s and 1s in certain input points where one input is a new value coming out from the current cycle and another input is the leftover "input-event" that esim produced in the previous cycle and kept it for the current cycle instead of iterating it in the previous cycle. Does anybody know a way or a hack to force esim to be more realistic ? I tried "s" "s" in succession and it did not work (the second "s" gave "simulation took 0 steps"). Are there are other simulators better than esim or built on top of esim? Can somebody give a strategy for simulating large circuits ? E. Sarmas topaz!sarmas