by Chandra Cheij

While some students use GPUs to frag digital foes, others use them to advance the frontiers of science (and perhaps one day unleash a cure for cancer).

In fact, thousands of students around the world are using GPUs today to tackle some of the most complex computing challenges in industries such as medical imaging, space exploration, automotive design and film production.

NVIDIA is involved in many academic research and educational programs, and one of the most important is the NVIDIA Graduate Fellowship Program. This program, which is open to students around the world, seeks to advance the frontiers of science by providing a financial reward and technical support to graduate students who are doing outstanding GPU-based research. 

In November, we kicked off the 10th Annual NVIDIA Graduate Fellowship Program and invited students to submit their program for consideration. Today I am happy to announce the winners of those Awards. Each student will receive $25,000, as well as engineering and technical support, to pursue their research.

Our Graduate Fellowship award winners were selected from over 200 applicants from 25 countries. Their projects involve a variety of technical challenges, including real-time high quality fluid simulations, a massively parallel MIMD accelerator architecture, motion planning, augmented operating systems, and programmability and optimization for heterogeneous systems.

Chris Malachowsky, co-founder of NVIDIA and senior vice president of research, said, “NVIDIA has a longstanding commitment to supporting outstanding academic research, because we know it fuels innovation. We’ve invested millions of dollars in support of ground-breaking research in such fields as 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.”

The NVIDIA Graduate Fellowship Program is open to applicants worldwide. Eligibility criteria includes 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.

The recipients of the 2011 NVIDIA Graduate Fellowship Program are:

Albert Sidelnik

Albert Sidelnik, age 31, from Edison, NJ
Studying at the University of Illinois at Urbana-Champaign

Albert is developing compiler techniques and language extensions to support high productivity and performance on exascale architectures that consist of hierarchies of accelerators and other parallel components.

“This funding is extremely beneficial in that it continues to allow me to work on cutting-edge research that will hopefully make it easier to use high-level languages for developing applications on next-generation parallel architectures. Additionally, this fellowship has helped me to connect and collaborate with other researchers and engineers who are tackling similar problems.”

Dominik Grewe

Dominik Grewe, age 24, from Germany
Studying at the University of Edinburgh

Dominik’s research is in the area of optimizing programs for heterogeneous multi-core architectures. Specifically he’s investigating ways to automatically partition parallel programs across processors in a heterogeneous system.

“In the Compiler and Architecture Design Group at the University of Edinburgh we are building up an excellence center for research on heterogeneous computing. The funding by NVIDIA will help me to further expand our infrastructure, allowing for more extensive experimental work in this area. Additionally the fellowship will allow me to visit other institutions and foster collaborations with research and industry.”

Etienne Vouga

Etienne Vouga, age 26, from Germany
Studying at Columbia University

Etienne’s project explores interesting physical simulations based on colliding objects. He seeks to develop an algorithm that can accurately and efficiently simulate contact.

“The NVIDIA funding will allow me to explore several exciting directions for the project. In collaboration with fellow student Samantha Ainsley, I expect I will be able to improve the performance of a preliminary implementation of the contact response algorithm by an order of magnitude, using ideas such as parallelization and cheap prediction followed by rewind and correction. I will also investigate applying the project’s ideas to the related problem of constrained motion, such as cloth that’s constrained to keep its shape and not stretch.”

Keon Jang

Keon Jang, from South Korea
Studying at KAIST

Keon’s work focuses on accelerating network processing such as IP routing and SSL, and building high-performance network equipment in a cost-effective way.

“The NVIDIA Graduate Fellowship will help me to focus on research without worrying about living expenses. Furthermore, I will attend conferences to broaden my knowledge and buy new GPUs to explore the research opportunities from a technology development.”

Michael Bauer, age 24 from Hereford, Maryland
Studying at Stanford University

Michael’s research is in the area of locality-aware programming models for deep memory hierarchies.  Michael works on both the design and implementation of languages and compilers that support this class of programming models.

“The NVIDIA graduate fellowship has provided me both tools and resources to substantially further my research by supporting additional hardware for multi-GPU experiments and enabling me to travel to new locales to describe the advantages of my research to new audiences.”

Michael Rubinstein

Michael Rubinstein, from Israel
Studying at Massachusetts Institute of Technology

Michael’s project focuses on utilizing parallel computing for long-range temporal analysis and modeling in video data.

“Long-range video analysis requires processing billions of pixels. NVIDIA’s cutting-edge parallel computing technology will allow me to analyze visual data in a much larger scale than is currently feasible for standard CPUs.”

Pinar Muyan-Ozcelik

Pinar Muyan-Ozcelik, from Turkey
Studying at the University of California, Davis

Pinar is investigating the use of GPUs in embedded systems that need to run multiple real-time tasks concurrently. Meeting embedded system constraints, providing real-time guarantees, and performing multitasking on the GPU are challenging research questions. In this study, she aims to find the best solution that can address these questions at the same time.

“This fellowship will support successful demonstration of my study, which will spur the greater adoption of powerful, inexpensive, low-end GPUs in embedded domains that require real-time multitasking. One of these domains is automotive computing, which requires concurrently running several data-parallel tasks such as speed-limit-sign recognition, e-mailing with speech recognition, infotainment applications, etc. In addition, this funding will allow investigation of the pros and cons of the current GPU programming paradigm for enabling multitasking of real-time embedded tasks, and consequently, will provide insights for developing next-generation GPUs and APIs.”

Sertac Karaman

Sertac Karaman, age 26, from Turkey
Studying at Massachusetts Institute of Technology

Sertac’s research effort explores massively-parallel anytime computation using sampling-based algorithms with a particular emphasis on the problem of robot motion planning. Applications include real-time planning for autonomous navigation of driverless cars in urban traffic as well as for unconventional aerial and all-terrain vehicles.

“This funding will allow us to explore the boundaries of massively-parallel real-time planning and decision-making algorithms for safety-critical robotic systems. Our hope is that this research will lead to deployment of parallel algorithms on real-world robotic systems to effectively solve challenging planning problems involving uncertainty, complex dynamic systems, and dynamic environments, a real-time solution to any of which is considered impossible today.”

Weibin Sun

Weibin Sun, age 27, from China
Studying at the University of Utah

Weibin’s project is to augment operating system kernels with GPUs. He seeks to run some OS kernel computation on GPUs to get higher performance and make it possible to integrate new functionality into the OS kernel.

“The NVIDIA fellowship will support my research by providing tuition, a GPU device for my experiments and a relationship with GPU experts at NVIDIA with whom I can consult. I also expect the fellowship to help me to get more attention, comments and feedback to my project from external professionals.”

Wen Zheng

Wen Zheng, age 28, from China
Studying at Stanford University

Wen is researching fast methods of physically-based simulation of natural phenomena through both algorithmic improvement and adapting to hardware architectures like GPUs with the goal of bringing Hollywood-quality visual effects to games and other interactive applications.

“This funding will pay for my tuition so that I can continue my research.”

Daniel Johnson, age 27, from Houston, TX
Studying at the University of Illinois at Urbana-Champaign

With the Rigel project, Daniel is investigating architectures for future accelerator processors with thousands of independent (MIMD) cores with the goal of enabling high throughput on a wide range of regular and irregular parallel computing problems.