GTC 2020 D2S30
Presenters: Tech Demo Team,NVIDIA
Simulations form an integral part of product design to reduce significant iterations in physical prototyping and testing to improve quality, cost and time-to-market. However, this process is very time consuming and can take weeks to months since, in a typical simulation workflow, several iterations are involved if the results are not satisfactory for a given design. Typically, there is never enough time or compute power to examine all the design variations.
The NVIDIA SimNet Toolkit is an end-to-end AI-driven simulation framework based on a novel Physics Informed Neural Network (PINN) architecture. This demonstration of SimNet is solving a multi-Physics problem to perform automatic design space exploration, a thousand times faster than traditional simulation, with the accuracy of numerical solvers.
Such unprecedented throughput enables optimized design selection, which we show using SimNet. Completing these design tasks take seconds not hours, and complex design optimization can be completed in days instead of months.
Watch this session
Join in the conversation below.