INFRASTRUCTURE FOR HIGH-PERFORMANCE COMPUTING
Dr. Apon, then a professor of computer science and Founding Director of the Arkansas High-Performance Computing Center, said that her work had been strongly supported by the state of Arkansas and by the NSF.40 The focus, she said, was cyberinfrastructure, which she said could be defined as the “IT infrastructure that enables scientific enquiry.” A statewide task force had developed a plan for cyberinfrastructure, with three entities:
The first was ARE-ON, the Arkansas Research and Educational Optical Network, an initiative to connect all the four-year public institutions in the state to a 10Gb network and provide access to state resources for anyone at any four-year institution. ARE-ON was scheduled to be fully operational at the end of the current semester, providing full access to national cyberinfrastructure resources. “The key message,” she said, “is that we are part of a national ecosystem.” She showed a list of collaborators in Texas, Louisiana, Pennsylvania, and elsewhere. “We can all share resources from our desktops at our four-year institutions.”
A Correlation between Federal Funding and Computational Capacity
Second, the Star of Arkansas was the state's largest computational resource, she said, funded through an NSF grant that provided about 11 million computing hours per year. “There is a lot of research in Arkansas that can benefit from computation,” she said. “One area is complex data analysis using emerging technologies to analyze data much more rapidly.” She added that there is an 80 percent correlation between a state's level of federal funding for computation and the state's ranking in computational capacity.
Another field in which computation is central is the accurate description of large molecules. She mentioned the work of Peter Pulay, a distinguished professor of chemistry at the University of Arkansas, Fayetteville, who studies the interaction of chemicals on human protein and DNA structure. Pulay's research requires 4 million hours of computing time each year.
Another computing challenge is found in nanotechnology, where scientists have already reached fundamental limits in computer technology. Research by Laurent Ballaiche has the potential to create nanotechnology devices that can build memory 10,000 times denser than anything currently manufactured, and his research requires 70 million hours of compute time per year. This is more than can be supported in Arkansas, she said, which is “why it's important to have access to national resources.”
In materials science, the challenge is to model plasticity and failure in metal alloys, with applications in aeronautics and other fields. Doug Spearot, assistant professor of mechanical engineering, creates 3-dimensional models of alloys that do not yet exist, using 20 million or more atoms. The computers evaluate variations of the alloys before they are fabricated in a laboratory. This modeling study requires 6 million hours of compute time a year.
The Most Important Instrument of Science
She emphasized the importance of supercomputing with a quote from Jay Boisseau, director of the Texas Advanced Computing Center: “Over the past 60 years, computing has become the most important general-purpose instrument of science.” Dr. Apon added, “My dad used to tell me, mathematics is the language of science. Well, computing is the most important general purpose instrument of science.”
In addition to ARE-ON, she said, NSF now funds a new EPSCoR Track 2 project called CI Train, or Cyberinfrastructure for Transformational Scientific Discovery. The intention is to provide campus cyberinfrastructure “champions” to serve as liaisons between the physical resources and the researchers and educators who need access to them. It also provides visualization resources and supports research in a wide spectrum of computational and visualization domains.
In Arkansas, she said, the CI Train project has made education a “key deliverable,” with partners across the state. It supports initiatives at the high school, undergraduate, and graduate levels with professional information technology staff and research faculty. It shared nationally competitive visualization resources and large-scale computational resources.
Re-thinking our Campus Environments
These new resources required unprecedented levels of sophistication for computer data visualization, she said. “We need to seriously rethink our campus environments and how they can support new data-driven modalities of research, collaboration, and education. She credited Rob Pennington of the NSF Office of Cyberinfrastructure for bringing this perspective, and also demonstrating how technologies can help scholars and students talk to each other and leverage all the resources across the state.
She closed with several examples from the NSF that pair problems with solutions. For example, recent computer science PhD training is disconnected from what scientists need; similarly, recent physics PhDs are not trained in software engineering. An example of a solution: create post-doc-to-professoriate programs to encourage them to apply their knowledge (and protect them).
“There are many needs,” she said in closing. “We must educate students at all levels in collaborative computational science. This is hard. Computer scientists don't necessarily want to do chemistry, and the chemists don't necessarily want to learn how to write programs on emerging technologies. It has to be a first-class, joint effort. We also have to encourage and support researchers moving into these areas, and we have to catalyze cultural changes in academics and agencies to better support interdisciplinary activities – not just by the faculty but also by the administration.”
RESEARCH PARKS IN ARKANSAS
“We are trying,” Mr. Chesshir began, “to take a state that has not necessarily been known for technology and innovation and move it into a brand-new world.”
With that, he said, he wanted to talk about research parks in Arkansas and what the state is doing to grow them. There were currently two science parks: the Arkansas Research and Technology Park, adjacent to the University of Arkansas in Fayetteville, and the Arkansas Bioscience Innovation and Commercial Center at Arkansas State University in Jonesboro, which is completing its Phase I business incubator. A new park was being constructed in central Arkansas.
The Arkansas Research and Technology Park (ARTP) used innovative techniques to nurture technology-intensive companies. It attempted to stimulate the formation of a collaborative community of companies, together with university faculty and students at Fayetteville, linked interdependently around a set of core R&D research competencies at the university.
Growing our own Expertise
“We learned in the last several years,” he said, “that people are not coming here in search of expertise. We're going to have to do a better job of growing it ourselves. We're not going to Boston, Ireland, or India to recruit that type of technology and innovation to come here and blossom. We are going to have to grow our own.”
The ARTP is an example, he said, of a community that has begun to create the next generation of electronic and photonic devices for biotechnology and related areas. These areas include transportation and logistics, in which Arkansas is a leader; materials and manufacturing; database software; telecommunications; and applied sustainability. Those are areas in which the ARTP is successful in terms of grants attracted and progress toward becoming a center of excellence.
The State's Primary Knowledge Community
A major advantage for the ARTP, he said, was its location in northwest Arkansas, near the main university campus. As the state's primary knowledge community, the Fayetteville area provided valuable fuel for the innovation and technology development activities of the ARTP. Two affiliates had received the prestigious Frost and Sullivan Award for excellence in technology, and another affiliate won the Tibbetts Award for the most innovative small business. Earlier in the year, another affiliate won an R&D 100 award, which cites Washington County as one of the most innovative in the country. ARTP affiliates, he said, continue to advance the frontier of product development in many specialty areas.
The reason that is important, he said, is that “what's going on up there in northwest Arkansas permeates the state, and provides a sense of innovation for folks in the other universities.”
In central Arkansas, a group had engaged a consultant to review activities at the University of Arkansas for Medical Sciences and the University of Arkansas at Little Rock. The question they asked was: How can we take the research and innovation that is already here and make it stronger? For so long, he said, the state had suffered from brain drain as its best and brightest young scientists, engineers, and medical researchers sought opportunities elsewhere. How, they asked, could the region take advantage of local innovative talent and turn it into jobs for the area and the state as a whole.
In 2007, this effort was rewarded when the General Assembly voted to create a research park authority, a legislative opportunity that would permit anyone in the state to create a research park and design it for sustainability. That effort had moved forward, he said, and at the end of the month, the authority was scheduled to be finalized with the city of Little Rock and its partners in central Arkansas, with the goal of beginning construction by 2012.
“This is a very ambitious schedule and investment,” he concluded. “It is something that has never been done in central Arkansas. Only when we have all our state, academic, and government partners working together are we going to be as successful as we need to be.”
UNDERSTANDING THE BATTELLE STUDY
Mr. Adams, who had founded the Arkansas Research Alliance after retiring from Acxiom Corporation, said that Battelle was asked to do a thorough study of economic development in Arkansas, primarily to provide evidence-based insight into the core competencies of the research universities. He said that at Acxiom, he had received many calls from people asking for funding for research that was unrelated to Acxiom's core activities. One lesson the company had learned, he said, was that it made sense to pay only for research that could move the country ahead. “So a key issue we discussed with Battelle,” he said, “was what are we good at in Arkansas? What can move us ahead?”
Battelle did a qualitative review based on field interviews with 85 of the top researchers in the state, and a quantitative review based on the journal publications and research grants of faculty members during the last five years. “In other words, by looking in the rear-view mirror.”41
Core Competencies and Economic Benefits
Nonetheless, he said, the study turned out to be a living document that revealed more than a dozen core competencies in Arkansas. But it also presented the challenge of finding the best way to derive more economic benefits from those 18 core competencies and 12 niche competencies.
“This was surprising to Battelle,” he said, “but 30 turned out to be too big a number. So we rolled them up into nine strategic focus areas.” Those were multi-disciplinary fields of research that were likely to enable the state to leapfrog more traditional universities that have more strength in narrow academic fields. The focus areas were also designed to engage multiple institutions, rather than be limited to individual universities or geography. The focus areas were:
- Enterprise systems computing,
- Distributed energy network systems,
- Optics and photonics,
- Nano-related materials and applications,
- Sustainable agriculture and bio-energy management,
- Food processing and safety,
- Personalized health research sciences,
- Behavioral research for chronic disease management, and
- Obesity and nutrition.
Increasing Multi-Campus Collaborations
“A reality of being a small southern state,” he said, “is that we're about $106 million below where we should be on a per capita basis in attracting federal research dollars to the state. Part of that is due to a lack of multi-campus collaboration, Battelle learned, so that a theme adopted by the ARA was to raise the number and level of multi-campus collaborations around those core competencies.”
The study looked at the competencies in terms of whether they were emerging, limited, or established, and examined the level of federal funding. “That is the accelerator,” he said. “State funding can help support talent, but federal funding is the key.” The next question was about market potential, the pull from industry, and whether there were already existing industries in Arkansas in these areas. He noted that Tom Dalton, of Innovate Arkansas, tried to “validate the technology: is this a business likely to stay in Arkansas, or will we create something that will move to Boston?”
He noted that the Arkansas Biosciences Institute already engaged in three of the core competencies, but that “we're not in competition with ABI, we're a partner with them.” He said that nine areas would eventually be too many to focus on, but that “we needed an evidence-based roadmap like Battelle's, as opposed to hearing researchers tell us how terrific their research is. Self-reported results have bias.”
A ‘Crucial Roadmap’ for Recruiting Talent
He said that the ARA had found the Battelle study to be a “crucial roadmap” to use in recruiting talent into the state, and into the core focus areas. For example, he said that Arkansas was in the process of launching an eminent scholars' program, modeled after one in Georgia.
The ARA was also trying to elevate the level of multi-campus collaboration, with funding from the Winthrop Rockefeller Foundation, the Walton Family Foundation, and the ARA board. With the help of administrative and research leaders of each of the five campuses, they had planned three conferences on (1) smart infrastructure, including the smart grid, (2) “smart information,” and (3) nanotechnology. They were also planning a conference on “healthy Arkansas,” and another based on bio-production and clean energy. The conferences would cover the nine focus areas during an 18-month period, with the goal of helping planners decide which to institutionalize in the state.
He closed by affirming that the “Battelle study has been a gift to this state. It's been absorbed into the EPSCoR conversation, and in the effort to focus the state's research resources. My hope is that we will use it as our investment roadmap going forward.”
Dr. Ahlen closed the symposium by exhorting his audience to go beyond the discussion stage and move into action. He observed that the Science and Technology Authority had a family of core programs and was managing two federal projects: an MEP project, with NIST, and an EPSCoR project, with NSF. The state and federal managers talked several times a week about these projects and other state activities. He asked, “How do we streamline these relationships?” NSF is primarily a research organization, he said, but it wants to see the results of research commercialized. “They tell the state we have to do that, and we do it through EPSCoR. But so do the state agencies that have been doing economic development for decades. The MEP would like to see more innovation in manufacturing. We applaud that, but we've also been trying to do that for 30 years.
“So it is time to look at these relationships, streamline them, and realize that we're all trying to move to the same place. We have multiple rules at the state and federal levels, and for those of us trying to execute, it's very difficult. All these rules are designed for transparency and accountability, but to different bosses in different places.
“I will remind some of us that 12 years ago, the National Science and Technology Council at the White House had its first interagency task force meeting on innovation partnerships, and after a couple of years the momentum came to a grinding halt. Here we are 12 years later having that same discussion. We don't have another decade to sit on this and wait for another discussion. We need to pick up the phone and call those friends in Washington who have told us to call.
“So,” he concluded, “go forward and collaborate.”
Dr. Apon is currently affiliated with the School of Computing at Clemson University.
Battelle Technology Partnership Practice, “Opportunities for Advancing Job-Creating Research in Arkansas, A Strategic Assessment of Arkansas University and Government Lab Research Base,” 2009. Access at http://www
.aralliance .org/__data/assets/pdf_file /0017/1682/Job-Creating-Research-in-Arkansas.pdf.
National Academies Press (US), Washington (DC)
National Research Council (US) Committee on Competing in the 21st Century: Best Practice in State and Regional Innovation Initiatives. Building the Arkansas Innovation Economy: Summary of a Symposium. Washington (DC): National Academies Press (US); 2012. Session V: Arkansas R&D Capacity: Universities Research Labs and Science Parks.