NVIDIA Announces Its Deep Learning for Science and Engineering Teaching Kit

NVIDIA has announced the creation of a Deep Learning for Science and Engineering Teaching Kit, which it says will help the next generation of engineers and scientists find out how to make best use of artificial intelligence (AI) technologies.

"We designed this course with my collaborator, Dr. Raj Shukla, to address the urgent need for specific material for scientists and engineers," says George Karniadakis, professor of applied mathematics and engineering at Brown University, who collaborated with NVIDIA on the project. "We focused on regression and mimicking the approximation theory and algorithms required in classical numerical analysis courses in the engineering curriculum."

The teaching kit is comprised of 15 lectures totaling more than 30 hours, 20 projects covering a range of fields, and coursework combining theory with practical examples — including, NVIDIA says, a primer on Python with libraries for scientific and deep learning computing, the training and optimization of deep neural network (DNN) architectures, physics-informed neural networks, neural operators, and, of course, high-performance computing with NVIDIA's own Modulus open-source framework.

"The NVIDIA Teaching Kit on physics-ML [physics-informed Machine Learning] has provided me with great resources for use in my machine learning course targeted for our engineering students," claims Hadi Meidani, associate professor of civil and environmental engineering at the University of Illinois Urbana-Champaign, who has had early access to the course. "The examples and code greatly enable hands-on learning experiences on how machine learning is applied to scientific and engineering problems."

The full course is accessible through the NVIDIA DLI Teaching Kit Program, while the lectures are openly published to the NVIDIA On-Demand platform.