News & Events
For Faculty & Staff
Clarkson University's Liu Leads Team on NSF Computer Cluster Project
Chen Liu, assistant professor of electrical and computer engineering at Clarkson University, has been awarded a three-year grant from the National Science Foundation to lead a four-person research team dedicated to acquiring a heterogeneous high-performance computing cluster.
Why a cluster? Sometimes researchers just need a super computer instead of the increasingly tiny versions that are common nowadays. This project is big, and it calls for big technology that can support parallel processing research of biometrics and identification technology, as well as broad disciplines of engineering research.
“If we are not the only ones, we are among very few in the country to have this ability. We are very happy and fortunate to get this grant,” Liu notes.
“This is a major research instrumentation grant, MRI for short,” he explains. “Our project is a small-scale super computer with a lot of horsepower for computation ability. It has many servers, interconnected to look like one big machine. Research involving facial recognition, iris recognition and fingerprint recognition requires a lot of computing power, so we're investigating how to perfect that capability and make biometrics run faster.”
A faculty member at Clarkson University for five years, Liu is the PI, or principal investigator, while his colleagues Professor & Chair of Mathematics Joseph Skufca and Paynter-Krigman Endowed Professor in Engineering Science Stephanie Schuckers are co-PIs. Clarkson's Chief Information Officer Joshua Fiske rounds out the team.
The project is challenging because in order to work with a very large database, such as faces, they need a lot of specialized hardware and software, and it needs to work quickly. Just for fun, the team will set up a demonstration machine on campus so passers-by can see their faces captured. Not to worry, their privacy will be protected, Liu notes.
The instrument is meant to support two very broad categories of research. On the one hand, application with a number of complex tasks operating at the interface of the physical and the computer world (cyber physical systems) requires rapid computations, but not all of those computations are of the same type. With a heterogeneous cluster, the diversity of the compute components allows for more efficient processing of these complex tasks. The face tracking and recognition problem is of this type.
The second type of research is more fundamental: if the first type of problem (the application) makes sense to tackle with a heterogeneous computer, how can we automate the parsing of the problem. Researchers will need to develop the algorithms that allow the computer to efficiently split up these problems so that the right kinds of tasks are sent to the right component in the cluster to maximize the capability of the overall system.
“A lot of people use computer clusters to do scientific research. Ours is a very unique concept," Liu adds. "It's called heterogeneous because there are four computing elements – central processing units (CPUs), graphics processing units (GPUs), many-integrated core (MIC) co-processors, and field-programmable gate arrays (FPGAs), tightly integrated with a light field camera as a data-capturing front-end."
Clarkson University is the lead site of the Center for Identification Technology Research (CITeR), a multi-university National Science Foundation Industry/University Cooperative Research Centers, working in partnership with government and industry stakeholders to advance the state of the art in human identification capabilities through coordinated university research.
The investigators plan to use this cluster to support research in CITeR, as well as other related areas. The team wants to involve the broader research community in this project. They can support other research across the faculty, Liu says, and Clarkson will support research at other universities that want to use this cluster.
“Our grant is for three years, but after the grant is done, the cluster will still be running,” he notes.
In making the award to Liu's project, the NSF notes that it “supports research, innovation, and discovery that provides the foundation for economic growth in this country. By advancing the frontiers of science and engineering, our nation can develop the knowledge and cutting-edge technologies needed to address the challenges we face today and will face in the future.”
This new cluster will be located in Clarkson's primary datacenter in the newly restored Old Main building. In addition to providing a home for campus computing resources, this facility also provides businesses in the North Country with convenient local access to an enterprise-grade datacenter. Through this new facility, Clarkson is offering access to a secure, stable and well-connected datacenter facility that can help growing organizations meet their technology infrastructure needs. For more information about the datacenter, visit http://oldmaincolo.com.
Clarkson University educates the leaders of the global economy. One in five alumni already leads as an owner, CEO, VP or equivalent senior executive of a company. With its main campus located in Potsdam, N.Y., and additional graduate program and research facilities in the Capital Region and Beacon, New York, Clarkson is a nationally recognized research university with signature areas of academic excellence and research directed toward the world's pressing issues. Through more than 50 rigorous programs of study in engineering, business, arts, education, sciences and the health professions, the entire learning-living community spans boundaries across disciplines, nations and cultures to build powers of observation, challenge the status quo, and connect discovery and innovation with enterprise.
Photo caption: Chen Liu, assistant professor of electrical and computer engineering at Clarkson University, has been awarded a three-year grant from the National Science Foundation to lead a four-person research team dedicated to acquiring a heterogeneous high-performance computing cluster.
[A photograph for media use is available at http://www.clarkson.edu/news/photos/cliu.jpg.]