Supercomputers, Superbuzzsaw

The four national supercomputer centers that were established last March have become the latest target for the Reagan Administration’s effort to restrict the flow of high technology to communist countries. The National Science Foundation will spend $200 million over the next five years to create the centers and make the superfast, superpowerful machines available at universities to a wide range of researchers in basic science. But the government, citing national security, wants those universities to agree to keep Soviet Bloc and Chinese scientists away from them.

So far the University of Illinois, Cornell University and a consortium of universities whose super-computer will be based near Princeton, N.J., have balked at signing an agreement that would restrict access to the machines. They argue that academic freedom would be compromised, that restrictions would clash with the ideal of open inquiry and that no one could learn anything from using the super-computers that couldn’t be learned elsewhere. Only the consortium based at the University of California at San Diego agreed to control access to its machine.

For several years the government has made the staunching of “technology leakage” a top priority, but it continues to go about it in the wrong way. Supercomputers have a potential military value in breaking codes and in designing weapons, but there is nothing inherently special about them that cannot be duplicated by using garden-variety computers in novel ways. Nor are the four super-computers in the new project the world’s only such machines. There are about 150 others in use, some in commercial operations, and the Soviets could presumably rent time on them if they wanted to.

In the past, when the Reagan Administration has attempted to restrict what foreign nationals may see and study on university campuses, the academic community held firm and forced the government to back down. The universities argued convincingly that progress in science depends on the free and open exchange of new knowledge. As a result, restrictions now apply only to classified information, which many universities refuse to undertake for precisely that reason.

Several years ago a distinguished panel of the National Academy of Sciences examined the government’s claims of injury to the nation from technology leakage and concluded that they were largely overstated. It recommended that high walls be erected around a few vital research areas and that the rest remain open.

The government appeared to be going along. But now the Department of Commerce has begun drafting regulations governing the exporting of genetic-engineering technology, and the new supercomputer centers have run into the national-security buzzsaw as well. The Reagan Administration needs to be constantly reminded that restrictions on research hurt the Americans’ efforts more than the Russians’.