[pseudosudoer]

There are search spaces that are quite large that are used in optimal control. GPUs can be used to drastically accelerate finding a solution.

As an example, imagine you are given a height map, a 2D discrete search space overlayed in the height map, 4 legs, and robot dynamics for every configuration of the legs in their constrained workspace. Find the optimal toe placement of the 4 legs. Although a GPU isn't designed exactly to deal with this sort of problem, if it's framed as a reduction problem it still significantly out performs a multi core CPU.

[whatever1]

These are sparse systems, factorization is not a strength of the gpu architectures. Typically adding more cpu cores is a better investment rather than trying to parallelize it through gpu. Nvidia has been trying for some time to make progress with cuSparse etc, although not much has been achieved in the space.

Maybe they try a completely different approach with reinforcemnt learning and a ton of parallel simulations?