DescriptionHerein, a novel algorithm for characterizing stress chains using a large parallel computer system is presented. Stress chains are important for analyzing the results of large-scale discrete element method (DEM) simulations. However, the general algorithm is difficult to parallelize especially when selecting networks longer than several particles. Therefore, we propose a new parallel algorithm to count the number of particles that are tightly connected, based on iterative operations with nearest-neighbor computations and communications. The new algorithm is examined via a real-scale numerical sandbox experiment using 2.4 billion particles. We successfully compute the stress chains with a reasonable computational cost comparable to the single-step DEM computation time. The visualization of the stress chains from the large-scale DEM simulation result reveals the existence of arcuate stress structures that may control accretionary prism formation, which is an important scientific discovery.