BEDRICK MAGAS is studying an unusually large gap in the ozone layer over Punta Arenas, a Chilean port set on the Straight of Magellan, almost at the very tip of South America. His research, at Universidad de Magallanes, could have much to say about changes in the global environment.
But partly because the university is located in a developing country, Dr. Magas doesn't have the funds and technology he says he needs for a complete project.
"Global-change studies, for example greenhouse-effect modeling and the ozone hole's impact on lower layers of the atmosphere, need massive computer power. Right now, the simulations are very short of accuracy because to date it has been impossible to include the effects on the vegetation cover in climate-modeling. You need a 1,000-fold increase in [available] computer power to know how the greenhouse [phenomenon] affects local [vegetation] in a 100- to 200-kilometer area," he says.
The kind of computer that could do this work is called a supercomputer. In the United States, there are almost 200 supercomputers, and they are used for everything from petroleum exploration to automobile materials design. There is only one supercomputer in Latin America - in Mexico.
A newly formed consortium of computer-equipment manufacturers and universities seeks to help fill this gap. Created a year ago, the Ibero-American Science and Technology Education Consortium (ISTEC) has already carried out student and faculty exchanges among universities in Latin America, Spain, and the US and is working to exchange and develop curriculum materials. (See related article.) Now the consortium wants to set up a supercomputer network linking the three continents, using satellite telecommunic ations.
Dubbed "Los Libertadores," or "The Liberators," the supercomputer project is meant to catapult Latin America into the next century, giving researchers access to the best technology available and to what others are doing in their fields.
"Now that all economic barriers are coming down, Latin America will have to compete on a level playing field with any industry in any place in the world," says Carlos Marino, a Colombian who is senior director of the Industry, Science and Technology Department at Minnesota-based Cray Research Inc., a supercomputer manufacturer.
Cray is designing the $40 million to $50 million, five-year Libertadores project and hopes to get funding for it from multilateral funding agencies, industry, foundations, governments, and consortium members. Motorola, Northern Telecom, Texas Instruments, John Fluke Manufacturing, and Sun Microsystems are also involved in the consortium.
Much of Latin America is decades behind developments in world technology because of scarcities in funding and trained personnel.
The experience of Ramiro Jordan, one of the consortium's founders, is typical of what happens when a Latin American wants to progress in his field. Born in Bolivia, he went to a university in Argentina and did his graduate work in the US. Today, he is an assistant professor of electrical engineering at the University of New Mexico (UNM).
"You're out there because you had the opportunity to make it," he says. "It's frustrating because you can't go back to where you were raised, where your family is, because there's nothing to do there, there are no job opportunities.... But you still have home in the back of your head and the feeling it would be nice to 'pay back' the country."
As world economic and political barriers topple, Latin America has moved away from government policies protecting local industry from imports. Many countries are actively seeking access to foreign technology, instead of closing themselves off to it. Last year, Dr. Jordan decided that the time had come for professionals like himself to involve themselves in the globalization process. Together with four other UNM professors, plus professors from Spain's Universidad de Vigo and the Universidad Politecnica d e Madrid, he founded ISTEC.
Members of the consortium met in November at Brazil's state-run Universidade de Campinas (Unicamp) to discuss the project and suggest new ways of exchanging and developing state-of-the-art technology for business use. In Latin America, academia has largely focused on basic research because it requires less equipment, leaving a gap between the academic community and local industry.
During Dr. Maris presentation on the Libertadores project at the meeting, one participant noted that, because of inadequate professional preparation, computer power at many Mexican universities is underused. Would a supercomputer really find the demand it merits, the participant wondered.
"When we hooked up the Mexican supercomputer to [our headquarters in] Minneapolis, we thought the Mexicans would be using about 40 hours a week of computer time. But they used 500 hours.... There is a lot of compressed demand," answered Marino.
Now, ISTEC is looking for funding for Los Libertadores and trying to decide which country will host the supercomputer and which others will have access via "mini-nodes." The consortium hopes to have the computing centers in place by 1993 and expects the project to become self-sustaining five years later.
Interviews with Latin American engineers found strong interest in activities of the consortium. Under its aegis, Unicamp and Northern Telecom are creating a joint research project.
Marisa de Giusti, a professor at Argentina's Universidad Nacional de la Plata, says international exchanges of professors will help to remedy her school's lack of a graduate engineering program.
With a supercomputer, "we could offer results and [industry] would trust us," says Julio Mendez, a Chilean chemical engineer trained at the Massachusetts Institute of Technology in Cambridge, Mass., and vice dean of engineering at Chile's Universidad de Concepcion. He adds that a supercomputer would find applications in Chile's copper, steel, petroleum, and paper industries, and could also be used in research on the country's air and water pollution problems.
Already, ISTEC has changed the lives of many students and professors, who have traveled over the last year to exchange ideas and work together. Unicamp graduate student Celso Jose Munaro spent three months visiting universities in the US.
"At Harvard, I talked with my 'idol,' engineering professor Roger Brockett," he exults. Mr. Munaro also visited MIT and spent two months at UNM, developing an idea for his doctoral thesis. Surprised to find how much applied research goes on in US engineering departments, he has changed his thinking on what kind of research is possible. "My dream is that someone can use my theory to build an intelligent machine, such as those programmable washing machines that sense color and the amount of dirt in clothin g," he adds.