NVIDIA Awards $275,000 to 11 GPU Computing Geniuses

by Chandra Cheij

We’re putting computing graduate research into high gear again with this year’s NVIDIA Graduate Fellowship Program.

The NVIDIA Graduate Fellowship Program awards $25,000 to Ph.D. students involved in computer research. The aim is to help students continue researching ways to use GPUs to tackle complex computing challenges in industries such as medical imaging, space exploration, automotive design and film production. The program provides a financial incentive and technical support to graduate students conducting outstanding GPU-based research.

In September, we kicked off the 12th Annual NVIDIA Graduate Fellowship Program and invited students to submit their research projects for consideration.

Our Graduate Fellowship award winners were selected from hundreds of applicants in 39 countries. Their projects involve a variety of technical challenges, including computer architecture, programming models, character animation, computer graphics and computational methods for simulating chemical events.

“NVIDIA is committed to supporting outstanding academic research because it fuels innovation,” said Bill Dally, chief scientist and senior vice president of research. “We’ve invested millions of dollars to support ground-breaking research in science, engineering and medicine. We’re delighted to support the work of these exceptional graduate students, whose efforts will help define the future of computing.” 

Recipients of the 2013 NVIDIA Graduate Fellowship Program are:

Zimmer HeadshotBrian Zimmer, from Berkeley, Calif.

Studying at the University of California, Berkeley

Brian’s research focuses on architecture and circuit-level techniques for improving energy efficiency.  He is developing circuits and systems that cope with the increasing variability of static random-access memory cells in deeply scaled technologies.


Wu HeadshotHaicheng Wu, from China
Studying at Georgia Institute of Technology
Haicheng is developing a compiler, Red Fox, for accelerating large-scale data warehousing applications on cloud architectures augmented with GPU accelerators. Red Fox is now capable of running all TPC-H queries in one GPU device with small-scale inputs.  The longer term goal of Red Fox is to be integrated with large relational database systems consisting of multiple nodes and multiple GPU devices to explore the opportunities for GPU computing in the “Big Data” era.


Hegarty HeadshotJames Hegarty, from St. Louis, Mo.
Studying at Stanford University
James’s research involves studying new programming models for CPUs and GPUs. He is examining image processing languages, with the goal of creating a programming model that is able to automatically exploit localities that are difficult for a general-purpose language to discover.


Fiss HeadshotJuliet Fiss, from Rochester, Minn.
Studying at the University of Washington

Juliet is studying computational photography, focused on light field video. She is working on the design of light field video cameras, light field image and video processing algorithms and light field video editing software that will allow filmmakers to control focus effects in post-production.


Luehr HeadshotNathan Luehr, from Alsip, Ill.
Studying at Stanford University

Nathan is a graduate student in the chemistry department at Stanford University. Working under Todd Martinez, Nathan develops computational methods for simulating chemical events. His particular research interests include efficient electronic structure algorithms for massively parallel architectures, and the application of molecular dynamics to large chemical systems such as proteins. 


Levine HeadshotSergey Levine, from Redmond, Wash.
Studying at Stanford University

Sergey is  developing learning algorithms that can allow virtual characters to emulate human behaviors. Such characters could be “programmed” simply by acting out the desired behaviors and recording the demonstration with a motion capture system. By allowing anyone to specify behaviors for virtual characters from demonstration, this technology could open up character animation to a much broader range of users, allowing more people to realize their creative ambitions and create engaging virtual characters. The same techniques could be extended for specifying behaviors for robots, or even studying motor control in humans and animals.


Tyree HeadshotStephen Tyree, from Tulsa, Okla.
Studying at Washington University in St. Louis

Stephen’s research specializes in machine learning on parallel computing hardware, rethinking and reengineering existing methods in light of current hardware trends.  He is porting support vector machine training to the GPU, and aims to contribute an open-source machine learning toolbox optimized for highly parallel platforms. 


Han HeadshotTianyi David Han, from Ontario, Canada
Studying at the University of Toronto

David’s primary research goal is to develop compiler and run-time support to make parallel and heterogeneous architectures easier to program. His current research focuses on optimization auto-tuning for GPU programs using machine learning techniques. This research builds upon his earlier work on GPU optimizations for reducing thread divergence and on hiCUDA, a directive-based interface that simplifies the process of porting a sequential program to CUDA.


Rogers HeadshotTimothy Rogers, from Halifax, Canada
Studying at the University of British Columbia

Tim’s research is focused on architectural changes to GPUs to improve their performance and power efficiency on highly parallel irregular applications, traditionally thought to be unsuitable for GPU acceleration. Such applications can be found in economically important areas like server and cloud computing.  Tim is focused on improving the way GPUs capture data locality when accelerating these workloads.


Chiang HeadshotWei-Fan Chiang, from Taiwan
Studying at the University of Utah

Wei-Fan is a Ph.D. student in the School of Computing at the University of Utah and co-advised by Prof. Ganesh Gopalakrishnan and Prof. Zvonimir Rakamaric. He received his Master’s degree from the University of Utah in 2010. His research interest is software verification in GPU or HPC context.


Lee HeadshotYunsup Lee, from South Korea
Studying at the University of California, Berkeley

Yunsup’s research explores ways to provide better performance and energy efficiency while retaining programmability and flexibility in data-parallel processors. He is developing new techniques to support irregular control flow more efficiently.


The NVIDIA Graduate Fellowship Program is open to applicants worldwide. Eligibility criteria include completion of the first year of Ph.D.-level studies in the areas of computer science, computer engineering, system architecture, electrical engineering or a related area. In addition, the student must hold a current membership on an active research team.

For more information on the NVIDIA Graduate Fellowship Program please visit our website.