IN THE DIGITAL DERBY, THERE'S NO INSIDE LANE

In the next 20 years, humankind will witness one of history's greatest technological transformations. At its heart are digital electronics. The trends are inescapable: The circuit density of computer chips--and hence their computing power--is doubling an average of every 18 months. Already, the next generation of $300 video-game machines, due out next year, will put in schoolkids' hands the graphics power of a 1980s-vintage Cray supercomputer. By 2010, the same device will fit in the palm of your hand and deliver photo-perfect images on a razor-thin display. By then, too, the transmission speeds of optical fibers will have jumped a hundredfold, enabling a strand of glass thinner than a human hair to carry a full-length movie in a fraction of a second.

At the same time, the unit cost of this computer power will decline dramatically. That will distribute discount digital knowhow to every corner of the earth, transforming corporations, redistributing wealth, and remodeling institutions. All this is happening at a far faster pace than the spread of technologies during the Industrial Revolution. In the digital age, product cycles are measured in months. Says Columbia University research fellow Bruce L. Egan: ``Technology has never moved so quickly.''

Or caused so much uncertainty. Over the next decade, world technology leadership will be up for grabs. Only a few years back, most pundits were sure that semiconductor manufacturing would migrate inexorably from America to Asia, particularly Japan. Today, the U.S. has more than its share of the newest, most advanced chip plants and rules the high end of the microprocessor market. And while many of Japan's most successful technology policies of decades past are getting clobbered in the 1990s, Silicon Valley still leads the world in digital technology, the key to breakthroughs in everything from multimedia to genetic engineering.

But new global challengers are emerging everywhere. Upstart South Korea has broken Japan's stranglehold on memory chips and consumer electronics. And since mathematics is the foundation of all digital advances, nations well versed in that discipline--including China, India, and the nations of Southeast Asia--could turn their homelands into formidable technology powers.

SMARTER SILICON. The extraordinary pace of change and the ever-denser web of global communications give the upstarts a big boost. Today, a breakthrough in a laboratory at Stanford University or Massachusetts Institute of Technology or a hot new device from Motorola Inc. can be rapidly designed into commodity products. As silicon gets ``smarter,'' even the most dazzling electronic products can be slapped together quickly in almost any location for less money than in the West or Japan. Unlike the infrastructure needed for the steel or petrochemical industries, which can take decades to transplant to a new country, the knowledge and software tools for designing new circuits can be taught in a classroom or downloaded over a telephone line.

Witness Taiwan. With an army of U.S.-trained talent, the country grabbed a huge chunk of the global PC assembly business in just five years. It took the Japanese 40 years to do the same thing in autos and a decade for memory chips. For national planners and chieftains at technology giants, this is a new world of big risks and baffling choices. Over the next 10 to 15 years, the entire information industry is likely to be restructured, although nobody can say just how. Will the world go wireless or be encased in optic fiber? Will the PC gobble up the functions of telephones, video games, and television, or will phone companies win that race?

To hedge their bets, companies are piecing together broad new technology portfolios. Often they are teaming up in joint ventures to shave the enormous R&D costs of staying ahead. Scaring them into action are the woeful tales of big bets gone wrong. Japan is saddled with unsalable high-definition televisions--relics of a multibillion-dollar bet on an analog technology that was eclipsed by digital video. Apple and AT&T can barely give away their ``personal digital assistants''--evidence of a misread tech road map. ``All the pathways of even thinking about industry will be fundamentally altered,'' says John Gage, director of Sun Microsystems Inc.'s Science Office in Mountain View, Calif.

The Information Superhighway will be a crucial agent of such change as hardware and software capabilities catch up with current market hype. Early in the 21st century, says AT&T Bell Laboratories President John S. Mayo, components based on light waves will start to augment slower electronic parts. The result will be seamless networks of data, voice, and moving pictures. In AT&T's vision, videoconferencing with built-in language translation across national borders will be as common as today's word-processing programs and spreadsheets. The technological building blocks for these advances are almost in place. ``The chief technology issue now is not `can we do it?' but `is it worth doing?''' says Mayo.

While AT&T plans the future, individuals will acquire more digital power of their own. This year, buyers around the world will snap up 44 million personal computers, according to International Data Corp. That includes 3.4 million running on Intel Corp.'s Pentium processors, whose power easily eclipses IBM's early mainframes.

CHANGING SOCIETY. Many will land in schools, hospitals, and government agencies yet untouched by the efficiencies inherent in computer technologies. Using high-speed videoconferencing links, specialists at some large U.S. medical centers, for example, can now diagnose and treat patients at remote locations--from small rural American hospitals to facilities overseas. The U.S. government is also pumping money into digital libraries that will give students around the country unlimited access to resource materials. ``The next stage of the information revolution will change not just industry but society itself,'' says AT&T's Mayo.

Twenty-first century digital horsepower is already at the fingertips of the global tech elite--executives like Sun's Gage, along with a few thousand top scientists and engineers in universities and corporate labs around the world. The computer at Gage's desk sends and receives data, sound, and digitized video at 622 million bits per second--43,000 times faster than what the typical modem plugged into a home computer can do. That speed allows Gage and his cohorts to swap detailed chip designs and run many scientific simulations, practically without a time delay.

Roll such systems out to millions around the globe, and the world economy changes profoundly. Retailers will contact customers on four continents. Shippers will track everything from container ships to parcels with centimeter precision using satellite-based global positioning systems. In publishing and entertainment, readers and viewers will pull news, movies, or documents directly from the ``bitstream,'' the vast galaxies of digitized video, sound, and data swirling in cyberspace. Even money goes digital, as greenbacks are replaced by electronic certificates with encrypted signatures.

Building the guts of this digital world--the hardware, software, fiber optics, and microprocessors--will realign the computer and information industries. In the world of multimedia and data superhighways, U.S. companies are in a stronger position than ever. Yet looking out over the next decade or so, Asia's new powerhouses will close the gap.

For a clue to how intense this conflict will become, take a look back at the drastic realignment of the world semiconductor and PC businesses over the past decade. From 1975 to 1985, U.S. engineers built dynamic random-access memory chips (DRAMS) into a multibillion dollar market and then ceded it to Japan. The Japanese planned to milk their hard-won monopoly long into the 21st century.

But their engineering skills and deep pockets were no match for South Korea's Samsung Group, which soared from ninth place in memory chips in 1989 to No.1 in 1993. Recently, Samsung was first out with a working prototype of the 256-megabit chips that will dominate the market by 2000. And market researcher Dataquest Inc. predicts that Samsung will be the world's top chipmaker of all kinds by 2005.

``INTELLECTUAL HUBS.'' Korea will not be the only Asian country vying for high-technology leadership. Once low-tech manufacturers of casings and keyboards, Taiwan's electronics companies now have two-thirds of the world market for personal computer motherboards--which contain the microprocessor and other chips. That success added mightily to the country's $90 billion in foreign reserves. Now, Taiwan is moving into networks and low-end telecommunications. Last year, a Taiwanese company called D-Link Corp. grabbed nearly 5% of the $1.4 billion U.S. market for so-called ``intelligent hubs,'' the circuit boards that enable PCs to swap information over local area networks. ``We didn't even track this company in 1992,'' says Dataquest analyst Marty Palka in San Jose, Calif.

The new Asian challengers will rely increasingly on commoditization--a process that combines American chip ingenuity and Asian product innovation. An early example was the generation of versatile U.S.-made chips that arrived in the 1970s. They allowed Japan's Casio, Sharp, and others to crank out millions of low-price digital watches and calculators.

In more advanced chips, such as those that control everything from cameras and mobile phones to auto engines, robots, and factory production lines, the U.S. is driving commoditization. These chips perform functions once handled by mechanical switches, motors, and other moving parts. By replacing these mechanical functions with mathematical formulas, called algorithms, embedded on silicon chips, companies such as Motorola, Intel, Texas Instruments, and LSI Logic deliver stunning price reductions. Since the chips can be fabricated in huge volumes, prices drop, both for the chips and the products that use them.

After a two-year, $15 million development project, C-Cube Microsystems, in Milpitas, Calif., unveiled a $25 compression chip that squeezes audio and video information so that a full-length movie will fit onto a single compact disk--a task that once required 10 separate chips. LSI Logic Corp. builds the powerful microprocessor that runs Sony Corp.'s new video-game machine, the Play Station. Now, LSI hopes to shake up the telecom world by bringing all the complex functions of phone switches and network controllers onto one piece of silicon. ``By the end of the decade, we'll be looking at 50 million transistors on a chip, and almost any electronic system will fit on it,'' says LSI CEO Wilfred J. Corrigan.

But there's a risk for the U.S. here, too. Nimble manufacturers overseas can buy these cheaper, more powerful American chips and software and ace U.S. competitors. In 1986, for example, Japanese auto makers turned to Silicon Valley to help them shorten their production cycles. Santa Clara-based Integrated Systems Inc. began providing automated design tools to Nissan and Honda. Using this software, embedded in tiny control chips, engineers could simulate a car's performance and modify the characteristics of prototypes while they were still on the test track. By not having to build a string of new prototypes, Japan shaved years off its production cycle. ``Detroit's attitude changed overnight,'' says Integrated Systems' Marketing Vice-President Moses Joseph. Today, General Motors Corp. and Ford Motor Co. are two of the company's largest accounts.

In high-volume electronics manufacturing, the risks can be more harrowing. Chips & Technologies Inc. in San Jose, for example, first created the low-price chipsets used to make motherboards. But the Taiwanese--with their manufacturing and low-wage cost advantage over American companies--quickly figured out a way to make the chipsets and motherboards more cheaply. They soon ran away with low-end PC manufacturing, including manufacture of some of the chips. Taiwan's next target: telecommunications.

But U.S. companies such as C-Cube Microsystems believe they have little alternative except to keep slashing costs and shrinking product cycles if they hope to stay on high tech's front lines. ``There's no choice but to run as fast as you possibly can,'' says C-Cube founder and Executive Vice-President Alexandre Balkanski.

Take a look at how Japan has stumbled. Companies such as Sony, Hi- tachi, and Toshiba spent 50 years building an empire of analog technology--televisions, VCRs, household appliances--that have since proved to be weak engines of profit growth. The companies also acquired superb digital skills, which they parlayed into profitable computer peripherals and audio/video gear. But much of the technology remained locked inside giant corporations, insulated from the software advances that have created America's most successful digital products.

That failure to go digital quickly has left Japanese companies even more vulnerable by creating an opening for a host of Asian competitors. ``Analog circuitry is an art that requires years and years of experience,'' says Lance Wu, deputy general director of Taiwan's governmental Computer & Communications Research Laboratories. ``That suited Japan's approach. But in the digital world, where speed to market counts,'' he says, ``Taiwan had the advantage.''

Moreover, Taiwan and other Asian challengers have benefited from a key human resource: the ethnic and intellectual bridge to Silicon Valley, sometimes called the Chinese Highway. Thousands of Chinese engineers--many with advanced degrees from U.S. universities--make up the largest ethnic minority in Silicon Valley. But there are many similar digital tribes crossing the global fields of high technology (page 164). Israelis dominate key parts of digital encryption, used in the new American cable- and satellite-TV equipment. Asian Indians run some of America's most prestigious research labs.

These groups are increasingly sought by technology giants for their expertise. A good example is the diaspora of Russian physicists and engineers uprooted by the collapse of the old Soviet research institutions. Last year, Samsung Electronics Co. demonstrated one of the world's first digital videodisk recorders, with far greater capacity than today's audio-CD systems. Samsung's secret weapon was blue-green laser technology licensed from Russian physicists. Sun Microsystems also recruited a group of Russians well grounded in a powerful chip architecture, which Sun will deploy in future workstations.

To remain successful, technology giants need to absorb a broad range of knowledge--be it from Asia, Russia, or the U.S. The huge costs of funding R&D, chip fabrication, and product development have forced many companies into strategic alliances. No company has exploited this strategy more shrewdly than Texas Instruments Inc. Since 1988, TI has initiated memory chip projects with Hitachi and Kobe Steel in Japan, Acer in Taiwan, and Canon and Hewlett-Packard in Singapore.

These alliances shelter TI from the cyclical slumps in the DRAM market. And in the upswings, the ventures generate cash that TI uses to diversify into higher-margin products, such as digital signal processors, whose booming sales fueled 11 consecutive quarters of profit growth.

Some strategic alliances require collaborating with a direct competitor. In the 1970s, Hewlett-Packard Co., hooked up with one of its fiercest rivals--Canon Inc.--to split a major investment in next-generation laser printers. Canon supplies the key mechanical component called the print engine, while HP provides the control chips and software. Outside this alliance, however, HP and Canon are locked in combat over a different technology--inkjet printers.

NOT FOOLPROOF. Canon's deal with HP may provide a model for a broader Japanese comeback. Nearly all of Canon's key alliances involve U.S. companies. It was the first company to license the PowerPC chip from IBM, for example, and has an equity stake in chip equipment leader Silicon Valley Group Inc. ``Seeking partners in Japan is futile,'' says Canon Senior Managing Director Hiroshi Tanaka. ``The Valley is 10 years, maybe 20 years, ahead of anywhere else. America is unmatched in every area of computer hardware and software.''

Even alliances aren't a foolproof hedge against the uncertainties of this vast digital transition, however. The stakes are highest in the battles over next-generation telecommunications infrastructure. Europe, America, and Japan are pumping billions of dollars into high-speed networks based on fiber and coaxial cable. Countries around the world installed more than 7.4 million miles of fiber last year, and the total will continue to grow 22% a year through the end of the century, predicts KMI Corp., a Newport (R.I.) market researcher. Estimates of the cost of rewiring America alone with fiber run as high as $400 billion, with private companies shouldering most of the costs.

But the wired world is just one option. Cheaper, wireless technology is also attracting massive investment. Developing countries could use wireless to bypass the wire-line phase. Already, digital phone and TV signals can be bounced off satellites, to be picked up by small dish antennas. BIG RETHINK. Such systems are ideal for delivering entertainment to millions of homes. In the U.S., GM Hughes Electronics Corp. and Hubbard Broadcasting Inc., which just launched a 150-channel satellite-TV service, project that there will be a half-million satellite-dish owners by yearend. By 1997, says CEO C. Michael Armstrong, Hughes will begin person-to-person videoconferencing services over high-speed satellite links.

This clash of alternatives--phone lines, cable, satellite, and cellular--is prompting a major strategic rethinking among phone and cable companies that are expected to pay for much of the Information Superhighway's cost. The confusion and the stalemate over regulation in Washington are stalling some critical experiments. During the past year, Tele-Communications Inc. backed off from a planned megamerger with Bell Atlantic, and few of the multimedia network trials initiated by US West, TCI, AT&T, and Time Warner are making progress.

The shakeup in the information industries is rippling all the way back to basic science. There, cheaper research methods promise to democratize what remains one of the most capital-intensive activities of the late 20th century. Supercomputers and sophisticated software have already reduced the need for expensively equipped biology labs or wind tunnels. Thanks to better instruments, more powerful simulations, and faster searches of scientific journals, ``we can do five times more in the labs than we did 15 years ago,'' says Joe A. Miller Jr., senior vice-president for R&D at DuPont Co. At their desktops, today's scientists and engineers have more sophisticated software than their predecessors who built atomic power plants and rockets in the 1960s.

Science has always been an international game. French biologists codisovered the AIDS virus, and the Max Planck Society has garnered over two dozen Nobel prizes in this century. Japan operates the world's largest atom-smashers for medical research. And the World Wide Web, a hot link on the global Internet, evolved out of the European Particle Physics Laboratory, called CERN.

BREAKTHROUGHS. But as science itself goes digital and supercomputer speeds migrate down to desktops, all geographical barriers are disintegrating, putting basic science within the reach of smaller and poorer nations. Scientists sitting in Taiwan's Hsinchu science park or in Seoul can run the same computational models as counterparts at MIT or Stanford. One result may be the erosion of America's commanding lead in basic research. Still, the gains or losses of individual companies or countries are trivial when viewed against the potential scientific breakthroughs ahead.

Indeed when so much brainpower is linked together, there are no permanent birthrights among individuals or nations. Despite jarring dislocations, the digital revolution has already distributed fantastically powerful tools across a huge sweep of humanity. That momentum will continue well into the 21st century.