U.S. Plan Could Give Computer Access to All

For all the current furor over drug wars, future decades may well remember the Bush Administration for an action it took Monday in proposing a $1.9-billion program to give as many Americans as possible access to the information bases of the most powerful supercomputers.

The Administration’s program, which would support research in computer software and hardware and create a National Research and Education Network, mirrors legislation already pending in Congress to expand Americans’ use of supercomputers.

The program ultimately could bring the whole of U.S. business and great masses of the American people into the information age, so that late in the coming decade Americans would flick on computers to tap the world’s knowledge as easily as they now flip a light switch.

That’s of historic and strategic importance, says Sheryl Handler, president of Thinking Machines Corp., a Cambridge, Mass., supercomputer maker. “Two hundred years ago, Napoleon said that God is on the side of the largest army. Today, he would just as accurately say that God is on the side of the largest supercomputers.”

Possible Competitive Edge

Supercomputers are those that can do more than 1 billion calculations per second and thus solve the biggest problems--how to organize missile defenses or map the stars or the human genetic system; how to find oil or design an airplane or an automobile.

But as those tasks imply, supercomputers are not the tools of everyday business. The Defense Department, giant corporations and the research centers of large universities use them, but the average company has no need to spend $20 million for extraordinary computing ability.

However, if an average company could affordably tap into supercomputing’s power and knowledge base to design its products or perform other tasks, the result could be a competitive edge. “The competitive effect is the difference between crawling on your knees and flying in a jet plane,” says Larry Smarr of the National Center of Supercomputing at the University of Illinois.

Which is why Japan is very intent on the supercomputer business, says Gary Smaby of Minneapolis’ Smaby Group, a supercomputer consulting firm. Years ago Japan mounted a national effort called the Fifth Generation computer project. It now has three companies--NEC, Fujitsu and Hitachi--making the big machines.

Encourages Creativity

And supercomputers have become bones of contention in trade talks, with the United States charging that Japan won’t buy American supercomputers for its universities--considered key customers because the engineers who train on your equipment may continue to favor it later in their working lives.

Still, U.S. industry leads the field, both in volume--with more of the world’s 400 supercomputers in place--and in technology. U.S. companies have devised numerous ways to solve such mysteries as the makeup of the earth’s ozone layer and the life of a molecule. “It’s really American reliance on creativity that has won so far,” says Handler of Thinking Machines--a 6-year-old company which has invented a computer that solves problems by using 64,000 separate processors simultaneously.

And creativity is the hope of the program announced Monday which is geared to expanding America’s use of the computer--building on a base of 30 million personal computers in use at present. The aim is to bring every university in the country into a supercomputer network and shortly after that, by the mid-1990s, to move to commercial services offering access to such repositories of knowledge as the Library of Congress and the National Centers for Disease Control. Users would be able not only to tap into such data but to choose quickly the pertinent facts they need.

Which sounds great. But if it’s such a good idea, why doesn’t business leap at the opportunity? Why do we need a $1.9-billion five-year government program?

Because the market right now is too diffuse, experts say.

An information network wouldn’t attract enough customers, and failure of a commercial effort could discourage further progress. “Technologies have withered from too few customers before,” says Elliott Maxwell, a vice president for policy issues at Pacific Telesis.

Ideal Government Program

The picture phone failed in the 1950s, he recalls. And though the jury is still out, videotext has not won commercial acceptance in the United States--although videotext networks are a great success in France where the government primed the pump by distributing personal computers to its citizens.

So a push from government is very much in order if the United States is to retain its lead in the information industries, says Senator Al Gore (D.-Tenn.) sponsor of legislation to create the National Research and Education Network. It would be an ideal government program, says Gore--one in which Washington would create a favorable environment and then step out of the way. The real objectives, from innovations in computing to national growth in the information-based economy, would be achieved by the users, the American people.