Higher Education and Research Conference Sessions

Hear how the University transitioned to AI, including  an overview of challenges and opportunities in curriculum changes, faculty transitions and infrastructure, research funding, and partnerships.

Discover the design strategy, diverse workloads, software stack, augmentations implemented, and lessons learned in deploying a hybrid GPU-based HPC cluster at the Milwaukee School of Engineering.

Learn how deep learning can be an effective weapon to defeat inequity in higher education through prior knowledge of students' performance. It can greatly help at-risk groups by offering aid to those who are more likely to struggle.

Learn how a state-of-the-art model for time-series classification was used for severe weather (tornadoes and severe hail thunderstorms) prediction. The model uses lightning data from the Geostationary Lightning Mapper (GLM) device aboard the NOAA GOES-16 satellite.

See how NVIDIA-based simulations complete the partial COVID spike protein structures, map the process of membrane fusion in atomic detail, and identify opportunities where drugs could block viral entry.

Explore how a  team from NVIDIA and the IceCube Neutrino Observatory at the Amundsen–Scott South Pole Station in Antarctica were able to accelerate the ray-tracing code use by the IceCube by 2x by porting it from OpenCL to NVIDIA(R) CUDA(R) and optimizing the algorithm.

We'll introduce educators to the Accelerated Data Science Teaching Kit, a comprehensive set of free teaching materials and hands-on programming assignments covering topics such as data collection and preprocessing, GPU-accelerated machine learning, and data visualization.

Learn about NVIDIA's Deep Learning Institute (DLI) University Ambassador Program, which provides educators the opportunity to get certified to teach instructor-led DLI workshops, at no cost. This can be done across campuses and academic conferences, covering topics like deep learning, accelerated computing, accelerated data science, and more.

Improve the sample-complexity and generalization abilities of reinforcement learning (RL) policies for robot manipulation tasks by incorporating object-centric priors in the RL agent's action space.

Learn about a new way to control a robot manipulator and control a robot's joints directly using a deep neural network.

Learn how to implement fundamental robotics in high schools with the Jetson Nano Developer Kit. Jetson’s AI curriculum teaches deep learning, python, and image processing algorithms.

We'll introduce educators to the Accelerated Data Science Teaching Kit, a comprehensive set of free teaching materials and hands-on programming assignments covering topics such as data collection and preprocessing, GPU-accelerated machine learning, and data visualization.

We'll introduce NVIDIA's Kaolin library for accelerating 3D deep learning research and demonstrate using the Omniverse Kaolin application to visualize training datasets, generate new ones, and to observe the progress of ongoing training of models.

Learn how to combine novel discontinuous, time-symmetric collocation methods with cuBLAS to accelerate the solution of Einstein’s equations in the time domain with double precision cores.

Find out about a collaboration with the Frontier Development Lab in partnership with NASA, the SETI Institute, and NVIDIA.

Learn the technical aspects related to the code developments and technology used (CUDA or OpenACC), the motives behind the approach, as well as benchmarking numbers on ChEESE use cases.

Discover how NVIDIA Jetson Nanos are transforming STEM programs for Kansas youth. See how one statewide youth development program is introducing youth to AI, robotics, and edge computing.

Learn about research on robot motion planning software that can optimize the speed and smoothness of robotic arm movement and reduce computation time for a pick-and-place task from 29 seconds to 80 milliseconds.

Find out about specific work on lidar-based mapping schemes that use panoramic depth maps as the underlying representation.

Discover the Duckietown platform, which provides a hands-on, accessible version of real-world autonomous systems.

Learn about a portable data parallel high-level interface for structured grid problems to GPU clusters and other architectures and review the detailed optimization of the code over both NVlink and Infiniband networks and for the A100 GPU.

Predicting side chain conformations of amino acids in proteins is both important and difficult. Learn how to tackle this problem using deep learning.

Learn about Tinker-HP  in the context of COVID-19 research, the performances and scalability of the multi-GPUs implementation as results from adaptive simulations using up to 1,000 GPUs simultaneously.

Find out more about simulating the flow around a silica-based sponge, the "Euplectella Aspergillum," using a TOP500 machine equipped with NVIDIA GPUs.

Learn about the development of GPU-accelerated meshfree solvers based on the Least Squares Kinetic Upwind Method for inviscid compressible flows.

Get an introduction on ExaWind, a high-fidelity modeling and simulation environment for analyzing wind turbine and wind farm flow physics.

Discover PyFR, an open-source Python-based framework for solving fluid flow problems using the FR approach.

Get an introduction to a graph neural network-based autoencoder model that provides effective reconstruction of detector simulation for LHC collisions.