The demise of the IBM PCjr personal computer after 14 months is a jarring example of the high mortality rates in the high-technology sector.

To some business analysts, the failure of the mighty International Business Machines Corp. in the home computer market, on the heels of similar failures by Atari, Mattel, Texas Instruments and Timex, demonstrates something very important about the nature of manufacturing today.

A combination of fierce competition and rapidly accelerating technologies has dramatically shortened the life cycles of products such as the home computer, compared with the typical life cycles for products in the 1960s and 1970s, these experts say. The PCjr's price was pushed so low by competition that IBM gave up on it.

A product is born, matures and dies in a matter of few years, leaving the manufacturer little time to recover development costs and profit from it: In some cases, the failure of a product line can take down an entire company -- such as Osborn Computer Corp.

Kenichi Ohmae, manager of the Tokyo office of McKinsey & Co., is one proponent of the theory. "Consider these time frames," he writes in his new book, "Triad Power."

"The first working computer was built in the 1920s. It took another 30 years to bring a digital computer from the development stage to a commercial product in 1954. . . . The integrated circuit, first developed in 1958 . . . took three years to become a viable product.

"Now consider the accelerated time frame for major developments in the semiconductor during the past decade." It took two years for U.S. manufacturers to increase the capacity of memory chips from 4,000-plus bits (units of information per chip) to 16,000 bits. But it took Japanese companies less than eight months to catch up. And the leap-frogging to higher and higher levels of memory capacity has been even faster.

"The rate of diffusion has become so fast that one can no longer assume a position of technological monopoly for long," said Ohmae.

That is the theory, and if broadly valid, it has critical implications for high-tech manufacturers, those that invest in them, work for them, and supply the components.

Such pressures would make large companies choosier about the products they bring to market -- resulting in more lost opportunities involving promising but risky technologies.

It would prompt more American companies to build or buy components offshore, or compel them to get into joint ventures with foreign competitors.

"I think it's a valid premise," says Richard M. Cyert, president of Carnegie Mellon University, one of the nation's leading centers of manufacturing technology.

The new technologies of computer-aided design and automated production lines are making it easier to design new products and bring them to the market, Cyert says. This in turn, is leading to shorter product life cycles.

"In science we talk about an inventor standing on the shoulders of earlier inventors," says Cyert. What's happening, he said, is that we're getting on each other's shoulders a lot faster than before.

The benefits are obvious. But the negative consequences "are fairly devastating," Cyert said.

IBM hasn't disclosed how much it invested in the PCjr, although the company is certainly big enough to handle the losses. On the human side of the ledger, the computers were made for IBM at a Teledyne Inc. plant in Lewisburg, Tenn., and substantial layoffs are expected among the work force of 1,600 employes there.

Setbacks such as the PCjr raise the risks and the hurdles to other high-tech investments, said Cyert, making companies "more and more short-run in their thinking. They're going to want to get a payback in two years instead of five, just when we ought to be thinking ahead 10 years.

"Companies right now have a choice -- to invest a large amount of money in a new production facility here or to buy what they need from Singapore or Thailand, he said. Not enough of them have the guts to build here, particularly in the face of a hostile economic climate. "And that's what's hurting the United States," he said.

Other experts argue that well-publicized failures like IBM's PCjr exaggerate the rate of change. "You can talk about accelerating change, but it's 25 years old" -- dating back to the invention of the integrated circuit, says Mitch Maidique, director of the Innovation and Entrepreneurship Institute at the University of Miami in Florida.

"I don't think the change is any faster than it was 10 years ago.

"The reason it looks faster now, is there are these guys 10,000 miles away who are giving us a tremendous run. Before we weren't worried about keeping up -- we set the pace. Now we have other people who are moving the pace . . . That is an incredible experience for us in this century."

But Bruce Merrifield, assistant secretary of Commerce for productivity, technology and innovation, agrees with Cyert. "I think it's pretty visible that the time cycles for given products are tending to compress.

"That means that the job of management is the management of continuous change. Those that fail to do that won't do very well," Merrifield said.

However, there is one positive consequence that that Merrifield and Ohmae both see -- an increasing collaboration among U.S. and foreign firms to share the costs of research and product development and then market a common product in several countries.

"It shares the risks, it pools resources and it opens markets that would otherwise be very difficult to penetrate," Merrifield says. And in the long run, it improves living standards and the quality of life around the globe.