A high performance algorithm that could solve problems with complicated materials on the consumer Gpus was developed by Chinese researchers, reaching a revolutionary increase of 800 times compared to traditional methods.
Developed by a research The team of Shenzhen Msu-Bit University, co-founded by Lomonosov Moscow State University and Beijing Institute of Technology, the new algorithm improves the computer efficiency of perishynatic (PD), a advanced and non-local theory that solves difficult physical problems such as difficult physical problems such as difficult physical problems such as difficult physical problems such as cracks, damage and fractures.
It opens up new possibilities to solve complex mechanical problems in various industries, including aerospace and military applications, on widely available fleas American sanctions.
The peridamic has proven advantageous in the modeling of material damage, but its high calculation complexity has traditionally rendered large -scale simulations ineffective, with problems such as high use of memory and slow treatment speeds.
To meet these challenges, Yang Yang, associate professor, expelled Nvidia Cuda programming technology To create the general PD framework. By carrying out an in -depth analysis of the unique structure of the chip, its team optimized the design of the algorithm and the management of the memory which led to a remarkable increase in performance. Their research was published in the Chinese Journal of Computational Mechanics on January 8.
“This effective computing power allows researchers to reduce calculations which would generally take days to a few hours – or even a few minutes – using an ordinary domestic level GPU, which is an important progression for MP research,” wrote Yang in the document.