NVIDIA has awarded $200,000 in Global Impact Award prizes to Princeton University and the University of Málaga for their pioneering research efforts to address social, humanitarian and environmental problems.
Researchers at Princeton’s Astrophysical Sciences Department have used deep learning to advance the future prospects for fusion energy feasibility, earning a $100,000 Global Impact Award.
Unleashing GPUs on a more than half-petabyte dataset, Princeton’s team, led by William Tang, has improved the predictions applied to experimental measurements from the world’s leading tokamak facility — the Joint European Torus.
Split-second predictions are necessary to head off disruptive events in multibillion-dollar thermonuclear fusion-grade plasmas.
Princeton’s researchers deployed their fusion recurrent neural nets deep learning software on 20 NVIDIA Tesla K20 GPUs to improve accuracy while reducing execution time to less than an hour, a task that previously took a day with CPUs.
Tang, a research professor in the Plasma Physics graduate program, will accept the Global Impact Award during NVIDIA’s GPU Technology Conference.
Read our blog post about the Princeton University team’s work.
University of Málaga Receives Global Impact Award
Spain’s University of Málaga received a $100,000 Global Impact Award for its work using GPUs to refine tsunami early warning systems.
Jorge Macías, an associate professor at the university and member of the Differential Equations group, is receiving the award, which recognizes the group’s novel GPU-based numerical model to accelerate tsunami simulations.
Know as Tsunami-HySEA, for simulations in the framework of tsunami early warning systems, the model’s ultimate goal is to prevent damage from future tsunamis and save lives.
Read our blog post about the University of Málaga team’s work.
Fourteen applications from half a dozen nations were submitted for the Global Impact Award for 2018. Other nominees this year were Massachusetts General Hospital, which is using deep learning in emergency rooms to prescreen faster for pneumothorax and other critical conditions, and the University of Washington, which is using deep learning to extend the usefulness of MRI machines.