Communications of the ACM
Nsf Grant Will Create 'big Data' Infrastructure at Ndsu
Credit: Big Data Insight Group
Research scientists are generating oceans of data, which can create challenges to capture, store, analyze, and understand. Standard computer systems cannot handle this so-called "big data" — high-volume, high-velocity data sets that are often in the terabyte and soon will be in the petabyte range. The U.S. National Science Foundation, in a competitive grant process, has awarded North Dakota State University a $400,000 grant over three years to create a Data-Intensive Cyberinfrastructure for Research and Education at NDSU, Fargo. The Office of the Provost will provide more than $171,000 in additional required matching funding. The computing infrastructure will be housed in NDSU's Research and Technology Park.
The award will enhance research capabilities at NDSU. It will also provide opportunities for high school and undergraduate students, as well as students from underrepresented groups in computational research, says Martin Ossowski, director of the Center for Computationally Assisted Science and Technology at NDSU, which will oversee the computational enhancements for researchers and a co-author of the grant proposal.
When completed, the system will be used by NDSU faculty, students, and staff for research in photophysics and photochemistry, plasma physics, new-generation energy conversion devices, multifunctional nanomaterials, biomimetics and coatings, computational biology, clay micromechanics, human brain injury under blast and impact, overland flow modeling, climatology and agroinformatics, structural monitoring of bridges, data mining, stochastic and bio-inspired computing, design automation of system-on-chip, and many other computational research areas. NDSU undergraduate students will be integrated into several of the research projects and data-intensive computing will be incorporated into senior design projects.
"The grant award will be used to develop and operate a new generation of advanced computing infrastructure at NDSU, Fargo," says Provost J. Bruce Rafert. "The resources that NSF provided to NDSU through this highly competitive award both recognize and accelerate NDSU's emerging leadership in cyber infrastructure." The new system will consist of tiered storage subsystems, tape library subsystem serving policy-driven near-line active archive, and a heterogeneous distributed memory compute cluster.
"These facilities will allow researchers access to additional state-of-the-art research computing resources, where 'big data' analytics are transparently coupled to high-performance modeling and simulation environments," Ossowski says. "What we are really excited about is that the system is designed to expand as NDSU's computational needs grow, by using what's called a 'condominium model' where individual researchers and research groups will be able to plug in their own hardware modules, resulting in unprecedented economies of scale."
The system will be tightly integrated with national resources including the National Science Foundation's Extreme Science and Engineering Discovery Environment (XSEDE) and the NSF-Department of Energy cross-agency Open Science Grid.
Outreach efforts of NDSU's Computational Research and Education initiative will include working with dynamic student groups. With the grant, NDSU will develop high performance computing (HPC) Summer Days to provide training in HPC for students in the Nurturing American Tribal Undergraduate Research and Education (NATURE) program. A cooperative effort between the state's research universities and tribal colleges located in the region, NATURE offers opportunities in science, technology, engineering, and math to tribal college students. NATURE is an initiative of tribal colleges and the North Dakota Experimental Program to Stimulate Competitive Research.
In addition, outreach efforts will include developing HPC Spring Days, a week-long program for local area high-school students. Proposed activities will include data intensive computing, as well as mentorship by an NDSU faculty member. Future outreach efforts will include HPC Fall Days, a project-based semester-long activity that pairs students involved in the ESTEEM Institute with NDSU mentors to tackle real-life computational science problems. The Institute, in its initial phases, includes area schools and colleges working to encourage opportunities for students in science, technology, engineering, and math.
The NDSU Major Research Instrumentation team will also work with Pavan Balaji, a computer scientist from Argonne National Laboratory. He serves as Chair of the Technical Committee on Scalable Computing (TCSC) of the Institute of Electrical and Electronics Engineers (IEEE). Balaji will augment outreach activities by using NDSU facilities as a computing test ground to conduct the Midwest portion of a data-intensive programming competition called the IEEE TCSC HPC Cup. NDSU's future outreach efforts also could include computational opportunities with private sector partners.
Twenty-five leading NDSU computational researchers from nine departments contributed to the proposal to the National Science Foundation. "The success of the proposal illustrates the importance of computational science as a unifying driver to researchers across the university," says Dinesh Katti, principal investigator for the successful proposal. "The rapid growth of computational power, along with important developments in computationally-driven science and engineering, has and will aid in major discoveries in a wide variety of fields."
"Computation often serves as a fourth dimension of research," says Philip Boudjouk, NDSU vice president for Research, Creative Activities, and Technology Transfer. "When looking for a needle in what are essentially billions of data haystacks, the new tools provided by this initiative become critical to researchers."
The Data-Intensive Cyberinfrastructure for Research and Education (DICRE) at NDSU will be managed by the Center for Computationally Assisted Science and Technology. The process to set up the computational infrastructure at NDSU is expected to take more than six months. DICRE at NDSU is funded by National Science Foundation Award No. 1229316.
No entries found