However, I'm not sure how much of a speed benefit it will give, for reasons discussed below.
The idea is to replace the asynchronous CA (henceforth ACA) with a stochastic synchronous CA (SCA) that behaves equivalently.
It's an interesting idea though, and I'll give some thought to it. Collision processing is effected entirely on the GPU is a step in a synchronous AC as shown in the link posted above.
for verification would use a local rule if Sum (neighbors) = 8 NO collision, Sum (neighbors) 8 Collision, it would be verified before running your update rule change if there is no collision cell states, as the two should be placed near the points to be evaluated if they are not close is that belong to other cells.
I should also clarify that an cellular automaton is a rather different thing from a synchronous one, and techniques for parallelising synchronous CAs (such as Conway's life) cannot easily be adapted to this problem.This takes $O(n)$ calculations, where $n$ is the number of grid cells.With some clever trickery, many of these can be done in parallel on the GPU side - but with the most obvious way of doing that you have no way of knowing where on the grid the collision was, so if there is one you have to start again.It's worth mentioning that I'm interested in the transient dynamics of these systems rather than just the equilibrium state, so I need something that has equivalent dynamics to the above, rather than just something that will approach the same equilibrium distribution.(So variations of the chequerboard algorithm are not what I'm looking for.) The main difficulty in parallelising the above algorithm is collisions.