The Discrete Element Method (DEM) has become a key computational mechanics approach for simulating particulate and granular systems. Recent progress in numerical methods, GPU-accelerated computing, contact model calibration, and multi-physics coupling has substantially increased the achievable system sizes, physical fidelity, and practical relevance of DEM simulations. This contribution reviews state-of-the-art DEM developments with an emphasis on numerical and computational aspects relevant to large-scale industrial simulations. Topics include efficient contact detection and integration schemes, treatment of non-spherical particles, CFD–DEM coupling strategies, and performance considerations for GPU-based implementations. Selected industrial use cases—primarily from pharmaceutical manufacturing processes such as granular blending, coating, and compaction—serve to illustrate how these numerical advances translate into improved predictive capability and process insight. The examples are demonstrated using the GPU-based DEM code XPS, enabling simulations of large particulate systems with complex physics at industrially relevant scales. The presentation further discusses modeling strategies that balance numerical accuracy and computational cost, as well as emerging directions such as AI-assisted parameter identification, reduced-order modeling, and digital-twin integration. Finally, key numerical and computational challenges for the broader adoption of DEM in large-scale engineering applications are outlined.