NVIDIA Awards $25,000 Fellowships to PhD Students for GPU Computing Research

by Sylvia Chanak

Our NVIDIA Graduate Fellowship Program awarded $25,000 this week to five Ph.D. students involved in GPU computing research.

“These exceptional grad students are helping to define the future of computing, and we’re delighted to support their work,” said NVIDIA Chief Scientist Bill Dally.

The fellowship supports graduate students doing GPU-based work. We selected our fellows from more than 100 applicants in 21 countries.

The winners:

chenfellowYu-Hsin Chen, Massachusetts Institute of Technology

Yu-Hsin researches energy-efficient computer vision frameworks for mobile platforms. He’s also looking into improving current GPU architectures with an emphasis on providing machine learning capabilities in multimedia systems.

cheungfellowBrian Cheung, University of California, Berkeley

Brian explores the hidden factors of variation learned in deep networks, finding correspondences between deep learning algorithms and the human visual system, and using data-driven methods to discover properties of visual processing in the brain.

heidefellowFelix Heide, University of British Columbia

Felix is interested in numerical optimization methods for large-scale inverse problems in computational imaging and vision. He uses GPUs to enable algorithms for statistically motivated image and vision reconstruction tasks such as non-line-of-sight imaging using time-of-flight sensors, where the goal is to “make the invisible visible.”

kellerfellowBenjamin Keller, University of California, Berkeley

Ben works on more energy-efficient digital design through optimization of the entire hardware stack, from architecture to devices. Smarter algorithms for dynamic voltage and frequency scaling, coupled with integrated on-chip regulators that achieve nanosecond switching and high efficiency, could reduce the energy consumption of many hardware applications.

pekhimenkofellowGennady Pekhimenko, Carnegie Mellon University

Gennady focuses on energy-efficient memory systems using hardware-based data compression. He discovered a series of mechanisms that exploit the redundancy in application data to perform efficient compression in caches and main memory, providing higher effective capacity and higher available bandwidth across the memory hierarchy.

This year, we additionally have a graduate fellow sponsored by our NVIDIA Foundation as part of its Compute the Cure initiative, which aims to advance the fight against cancer: 

neylonJohn Neylon, University of California, Los Angeles

John researches adaptive radiation therapy-based cancer treatments. He’s developing a framework using image registration and predictive biomechanical models for regression tracking, dose estimation and extrapolation. Accelerating these tasks with GPUs will allow integration into existing clinical workflows and provide physicians with information for optimizing treatments to the patient’s anatomy.

We also acknowledge our six finalists:

  • Neha Agarwal, University of Michigan
  • Forrest Iandola, University of California, Berkeley
  • Viktor Kampe, Chalmers University of Technology
  • Ji Kim, Cornell University
  • Jui-Hsien Wang, Cornell University
  • Liwei Wang, University of Illinois, Urbana-Champaign

The NVIDIA Graduate Fellowship Program is open to applicants worldwide.