You are currently viewing NVIDIA Pronounces Its Deep Studying for Science and Engineering Instructing Equipment

NVIDIA Pronounces Its Deep Studying for Science and Engineering Instructing Equipment

NVIDIA has introduced the creation of a Deep Studying for Science and Engineering Instructing Equipment, which it says will assist the subsequent technology of engineers and scientists learn how to make finest use of synthetic intelligence (AI) applied sciences.

“We designed this course with my collaborator, Dr. Raj Shukla, to deal with the pressing want for particular materials for scientists and engineers,” says George Karniadakis, professor of utilized arithmetic and engineering at Brown College, who collaborated with NVIDIA on the venture. “We targeted on regression and mimicking the approximation idea and algorithms required in classical numerical evaluation programs within the engineering curriculum.”

The educating package is comprised of 15 lectures totaling greater than 30 hours, 20 tasks masking a spread of fields, and coursework combining idea with sensible examples — together with, NVIDIA says, a primer on Python with libraries for scientific and deep studying computing, the coaching and optimization of deep neural community (DNN) architectures, physics-informed neural networks, neural operators, and, after all, high-performance computing with NVIDIA’s personal Modulus open-source framework.

“The NVIDIA Instructing Equipment on physics-ML [physics-informed Machine Learning] has supplied me with nice sources to be used in my machine studying course focused for our engineering college students,” claims Hadi Meidani, affiliate professor of civil and environmental engineering on the College of Illinois Urbana-Champaign, who has had early entry to the course. “The examples and code vastly allow hands-on studying experiences on how machine studying is utilized to scientific and engineering issues.”

The complete course is accessible by way of the NVIDIA DLI Instructing Equipment Program, whereas the lectures are overtly printed to the NVIDIA On-Demand platform.

Leave a Reply