Almost two years ago, we decided to put COMP Performance Group’s DynoSim 5 software to the test. Everyone has always hemmed and hawed at whether such software is accurate, but the overarching sentiment around the office was general skepticism. So we set out to see for ourselves.

Since the old computer science term is “garbage in, garbage out,” we put a whole lot of time and effort into doing things right, so that whatever the results were, they were honest and repeatable. We even went as far as to enlist one of the smartest guys we know when it comes to the valvetrain, COMP Cams’ Valvetrain Engineering Group Manager, Billy Godbold.

As a baseline, and to act as a control group, we modeled the engines from the original LS vs. Coyote Shootout series, since we not only knew every single part that went into each engine, but we had detailed flow data on each cylinder head used along with comprehensive dyno results.

After “running” the modeled engines, we found the simulation to come in at 3.5-percent high on the horsepower numbers, and plus-or-minus 1.5-percent on the torque numbers, with the dyno curve shapes and peaks matching the real-world data exactly. That gave us the confidence to try something that the critics of the first competition were very vocal about — matching displacements.

The two modeled dyno curves from the two validation engines (LS3 is the lighter red and lighter green lines), both of which match the shape of the real-life dyno curves. Both hypothetical engines came in at 3.5-percent high on the power numbers, and within 1.5-percent on the torque numbers.

To do that, we took a hypothetical 4.8-liter LS engine, added a little bore to bring the displacement to 5.0-liters, added some compression to equal that of the Coyote engine used in the original test, selected the induction parts we used in the validation (which are the same as what we’d use in real life), and chose an off the shelf COMP camshaft (after running significant testing for improvements using the software).

The result of all this was something no one really expected, and that was an LS engine with a very similar power curve to the Coyote engine, and only falling off in the top end, presumably finding the limits of the stock 799 heads modeled. Besides being an extremely fun article to write, we also found a reliable took in the DynoSim5 software.

It has played a large part in the pre-planning stages of several projects around the office, as well as inspired several others that are currently just “on paper.” However, if you ever see a 5.0-liter LS engine project pop up, it’s a safe bet you know where the idea came from.

The final 5.0-versus-5.0 comparison graph. The lighter colored lines are the LS, and the darker colored ones are the Coyote engine. I’m not sure anyone involved in this article expected the curves to be this similar, and the overall numbers to be so close to one another.