Peridynamics (PD) demonstrates unique advantages in addressing fracture problems, however, its nonlocality and meshfree discretization result in high computational and storage costs. Moreover, in its engineering applications, the computational scale of classical GPU parallel schemes is often limited by the finite graphics memory of GPU devices. In the present study, we develop an efficient particle information management strategy based on the cell-linked list method and on this basis propose a subdomain-based GPU parallel scheme, which exhibits outstanding acceleration performance in specific compute kernels while significantly reducing graphics memory usage. Compared to the classical parallel scheme, the cell-linked list method facilitates efficient management of particle information within subdomains, enabling the proposed parallel scheme to effectively reduce graphics memory usage by optimizing the size and number of subdomains while significantly improving the speed of neighbor search. As demonstrated in PD examples, the proposed parallel scheme enhances the neighbor search efficiency dramatically and achieves a significant speedup relative to serial programs. For instance, without considering the time of data transmission, the proposed scheme achieves a remarkable speedup of nearly 1076.8× in one test case, due to its excellent computational efficiency in the neighbor search. Additionally, for 2D and 3D PD models with tens of millions of particles, the graphics memory usage can be reduced up to 83.6% and 85.9%, respectively. Therefore, this subdomain-based GPU parallel scheme effectively avoids graphics memory shortages while significantly improving the computational efficiency, providing new insights into studying more complex large-scale problems.

Boron carbide, a typical lightweight and high-strength ceramic material, has broad application prospects in national defense, military, and aerospace. However, the nanoscale amorphous shear band, which mainly arises from the destruction of icosahedra, is easily formed in boron carbide under impact, thereby causing its catastrophic shear failure. Since the formation of amorphous shear band of boron carbide significantly depends on its microstructures, molecular dynamics simulations have become a key approach to reveal the microstructural evolutions and mechanisms. However, due to the insufficient accuracy of classical atomistic potentials, classical molecular dynamics simulations face significant challenges in simulating complex material systems, such as boron carbide. In recent years, the development of machine learning methods has provided a new research paradigm for the development of atomic potentials. Among numerous machine-learning atomistic potentials, the deep potential (DP) model, which is based on deep neural networks, is particularly widely applied. This DP model can not only maintain the accuracy comparable to that of ab initio simulations, but also exhibits the efficiency comparable to that of classical molecular dynamics simulations. Thus, the DP model has become an effective strategy to examine complex material systems. In the present study, we systematically examine the research of the DP method on boron carbide ceramics. Firstly, the theoretical framework, development process of the DP model, and the construction and validation of the DP model for boron carbide are summarized. Subsequently, the mechanical responses and the localized amorphization mechanisms of boron carbide are revealed using deep potential molecular dynamics simulations. Then, some strategies are proposed to enhance the ductility of boron carbide, including microalloying, stoichiometry regulation, grain boundary engineering, and defect control. Finally, the application prospects of the DP model in the research of complex material systems, such as boron carbide, are explored.