Chinese scientists, in collaboration with Russians, have developed a computing algorithm that dramatically improves the performance of NVIDIA graphics cards in scientific and industrial calculations.
Researchers at Shenzhen University MSU-BIT and the Beijing Institute of Technology, in collaboration with Russian scientists (Moscow State University), have developed an algorithm that can significantly improve the efficiency of peridynamics calculations — a nonlocal theory used to model the strength and fracture of materials. It can speed up the calculation of material simulations by up to 800 times.
Peridynamics is widely used to predict material failure in aerospace, civil engineering, and military applications. Its traditional algorithms require significant computing resources, making large-scale research slow and impractical. Associate Professor Yang Yang and her team solved this problem by using NVIDIA’s CUDA technology to optimize algorithm design and memory management.
Their PD-General framework achieved an 800-fold increase in computational speed for NVIDIA RTX 4070 video cards compared to traditional sequential programs and 100x performance improvement compared to OpenMP parallel programs. In a large-scale simulation with millions of particles, PD-General completed 4000 iterative steps in five minutes. For large-scale two-dimensional uniaxial tensile problems, it processed 69.85 million iterations in less than two minutes with single-axis accuracy.
The increased computational efficiency means that researchers can now run simulations on consumer-level graphics cards instead of relying on expensive high-performance computing clusters. This has broad implications for industries that require detailed material analysis, including military and aerospace, construction and industrial development, and more.
The ability to run high-performance modeling on affordable graphics cards will help the Chinese and Russians circumvent sanctions imposed on the supply of powerful computers Western countries. For Ukraine, this means improving Russia’s ability to develop military technologies.
The study was published in the Chinese Journal of Computational Mechanics in January 2025. The researchers believe that further development will help improve calculations in other scientific and industrial applications.
Sources: SCMP, Tom`s Hadware
Spelling error report
The following text will be sent to our editors: