Here’s a paper where we used ADMM to parallelize (and greatly speed up) the computation of solutions for a class of stochastic games.
This work was done in collaboration with Jingqi Li, Frank Chiu, Somayeh Sojoudi1, Claire Tomlin at UC Berkeley, Lasse Peters and Javier Alonso-Mora at TU Delft, and my colleague Mustafa Karabag and advisor David Fridovich-Keil at UT Austin.
We presented this work at the 62nd IEEE Conference on Decision and Control (CDC 2023) in Singapore.
Find the code here.