Big Award By Army Sparks Debate On Control Of Academic Research

University of Minnesota math chairman resigns in controversy after his school is awarded a $67 million computer center contract MINNEAPOLIS -- Six months after the successful end to a heated competition to win a $67 million supercomputer research grant from the Army, University of Minnesota computer scientist Ettore Infante is trying to track down a bumper sticker portraying him as a warmonger. "It shows me on the head of a missile," says the genial dean of the university's Institute of Technology. "My hair is bent back with sparks coming off. I hear it's something. And I'm holding a lit bomb." Infante has been unable to find the bumper sticker, reportedly put out by a student group protesting the Army contract. And he's not much concerned that the sentiment behind it is shared by many members of the institute's mathematics department. Infante is convinced the Army contract, the largest single grant ever awarded to an institution in Minnesota, is worth the rebellion it has engendered among students and faculty. The protests include the resignation of Richard McGehee, chairman of the mathematics department, who described Infante's rush for the Army money as "disgusting." The contract, won by Minnesota last August over nine other universities, establishes an Army High Performance Computing Research Center (AHPCRC) at the university-affiliated Minnesota Supercomputer Center here. The supercomputer center, a private corporation partially owned by the university, will donate to the project about 1,400 hours of time per year on its Cray 2 and Cray X-MP supercomputers - time worth $5 million over the five-year life of the contract, according to Infante. The University of Minnesota is the lead school in the project, which also includes Jackson (Miss.) State University, Howard University in Washington, D.C., and Purdue University in West Lafayette, Ind. In addition to participating in the project with visiting faculty members, Jackson State will help to recruit minority graduate students. Howard University, which is developing a strong program in computational mathematics, will be a "major participant" in the project and will receive about 10 percent of the grant money, Infante says, while Purdue's role will be relatively minor. Another member of the consortium is the Computer Science Corp., based in Washington, D.C. The controversial center is designed to increase the Army's ability to use supercomputers for such things as battle management, simulating how armor and newer weapons materials respond to heat and stress during combat, and identifying friendly and hostile troops through the use of radar and other battlefield systems. While that may be how the Army uses the knowledge gained from the project, Infante says, university faculty won't participate in any classified work. Instead, their research will focus on the challenges inherent in linking the powerful memory of its Cray computers with parallel processing computers such as the Connection Machine made by Thinking Machines Corp. in Cambridge, Mass. A decision on what parallel computer to buy has not been made, he says. "The biggest Cray has eight processors and the Thinking Machine has 64,000," he says. "If you think of computation as a cart, is it better to pull it with four horses [the Crays] or 1,064 chickens [the parallel machines]?" he asks. "That's the long-standing argument. It's much easier to control the four horses, but the fine-grain parallel machines have more computational power. We want the four horses and 1,064 chickens to run together." About $27.3 million from the grant will go to purchase parallel machines and other computer equipment. Another $8.5 million is designated for research. The largest single block of money, about $31 million, falls under the general category of support services and includes $5 million to the university for administrative support, $14 million to the supercomputer center, and $11 million for training Army computer specialists. Objections to the project by the mathematics faculty were strong. When word reached them that Infante was planning to submit an application, 50 of the department's 75 members signed a letter to university officials objecting to the way the bid was being put together. "That [letter] was ignored," McGehee says. "There was an awful lot of hiding and secrecy." The objections among the math faculty centered on two issues, he says. "A number of people don't believe the Defense Department should be sponsoring pure scientific research," McGehee says. "That's not the majority of the 50 who signed the letter, but there are many who feel that way." Another aspect of the contract that bothered faculty was the possibility of classified research eventually being done at a university that has a policy against it. Donald Aronson, a math professor who strongly opposes the project, is concerned that the Minnesota Supercomputer Center could circumvent the uni-versity's ban on classified research because it's a private corporation. But Infante says no faculty members will conduct classified research, and he doubts if university researchers would even "have clearance to classified material." In a memo last June to university officials, Infante said classified work would be done at Army laboratories and at the Computer Science Corp. While the military nature of the contract resulted in demonstrations and disruption of regents meetings by some students, the mathematics department was even more concerned about the scientific issues it raises. "Who sets the priorities for the university?" asks McGehee. As part of the contract, for example, the university promises to spend $750,000 per year of its own money to hire new faculty members, in several departments, with the skills to aid the project. Infante says eight new faculty members are being hired to "increase the talent we have, especially in computational mathematics and engineering and computer science." But McGehee objects to the process that was followed. "These decisions were made without consulting the faculty, who should be the ones who decide which direction we go in." Traditionally, McGehee says, bids for outside grants are prepared by faculty members and must work their way up through the university bureaucracy. The procedure includes approval of all affected department heads. But the Army contract "was Infante's idea," McGehee says, "and he had trouble finding people who would handle it. The proposal wasn't available for us to see, and it was submitted without any signatures [from the math department]." Infante says the process was proper and faculty members are upset only because "they weren't given the chance to say no, no, no. When we submitted our best and final proposal we put it under the noses of all the department heads, and everybody signed but Mc-Gehee." McGehee's resignation is effective at the end of the academic year, and he will remain as a professor in the math department. He says he'll watch the work at the center very closely. "I raised quite a few objections and was ignored," he says. "I don't believe in continuing the battle, and I don't see launching a campaign to get rid of the center. I'll try to the extent possible to make sure the money gets spent well." McGehee says colleagues at other universities that bid on the project are relieved that the contract went to Minnesota. "I got the impression that a lot of universities handled [their proposals] the same way ours did," he says. Mathematicians at a number of schools "breathed a sigh of relief" when their schools failed to land the contract, he says. Universities that competed for the contract included the University of California, San Diego. Mathematics professor George Sell has been appointed director of the project, but Infante remains its spiritual leader. "The Army people call me the Godfather," Infante says, because of his Italian name. Asked if he found that offensive, Infante laughs. "When it comes to $67 million," he says, "they can insult me all they want." Jim Dawson is a science writer with the Minneapolis Star-Tribune.

Big Award By Army Sparks Debate On Control Of Academic Research

Interested in reading more?

Become a Member of