AI-Accelerated Computational Science and Engineering Using Physics-Based Neural Networks
, NVIDIA
AI has great potential to address the performance gaps in traditional simulations for science and engineering problems without sacrificing accuracy. The area of application is very diverse — from real-time simulation (e.g., digital twin and autonomous machines) to design space exploration (generative design and product design optimization), inverse problems (e.g., medical imaging, full wave inversion in oil and gas exploration) and improved science (e.g., micromechanics, turbulence). We'll cover state-of-the-art AI for addressing the above-mentioned applications that are difficult to solve because of various gradients and discontinuities, due to physics laws and complex shapes. For desired convergence and accuracy, we'll discuss development of new network architectures for such problems and implementation of special techniques specific for physics and geometry, as well as large-scale parallelism and GPU architecture-specific enhancements for good performance.