Hi Nathaniel,
I am performing these simulations through a Revit plugin called "Aftabrad" so I am not sure if you would be familiar with it, however the files it outputs to perform SDA calculations are
linked here.
There is also an image there with the rtrace parameters used in the simulation.
For the results essentially I am getting consistent results with minimal difference when changing the ambient parameters and limit weight (I have tried 1/ad, 1/ad^2). However when compared to results with the GPU off the accelerad results have higher values. This could possibly be an issue due to how Aftabrad performs calculations so I am not entirely convinced it is an accelerad issue.
Thanks,