The Center for Scientific Communication of the Moscow Institute of Physics and Technology reported that a new technology based on machine-learned potentials will help use AI to create high-precision models of materials and production processes, including welding and heat treatment.
Machine-learned potentials are specialized AI systems trained on limited amounts of full quantum chemical calculation results. They can spend three times less time analyzing information than traditional algorithms.
To optimize the performance of these AI systems, an efficient implementation for GPU should be developed, as well as the addition of long-range interactions and improved methodologies for creating databases for modeling. In addition, the integration of experimental data promises further improvement in the accuracy and applicability of machine-learned interatomic potentials.
Now scientists from different countries are wondering how to use these algorithms, and have identified the pros and cons of such an application of already existing machine-learned potentials.
Researchers from Russia presented calculation systems based on potentials of moments of inertia (MTP). In terms of accuracy, MTP is on par with global counterparts, in terms of performance even surpass them. And the main difference is that Russian calculation systems can be trained on fragments of complex systems.
So far, these technologies cannot be used on video cards (GPU). But scientists are on the way to discovering this possibility, because GPUs have higher computing power than CPUs (central processors), where the method of potentials of moments of inertia is still used.
Related materials:
Scientists from Chelyabinsk will create the first quantum infrared microscope in Russia
AI technologies will help develop the infrastructure of 6G networks
