AT&T; BREAKUP II : OUTLOOK: THE COMPUTER COMPANY : Spinoff of Unit Could Be Last Hope in PC Industry : Technology: After struggling to gain market share, the new focus will be on previously successful specialty markets.

AT&T;’s spinoff of its Global Information Solutions group looks to be the phone giant’s best perhaps only--hope in the computer industry, a market where success has proven elusive despite the billions of dollars the company has spent over 10 years.

“It’s no secret that our computer business . . . is going through a difficult transition,” AT&T; Chairman Robert Allen said Wednesday. “But is it a lost enterprise? Absolutely not.”

At the same time that it announced its intention to split into three companies, AT&T; laid out plans to resuscitate its flagging computer business, which lost $322 million on revenue of $3.9 billion during the first six months of this year. Global Information Services will lay off 8,500 of its 43,000 employees, cutting away some of the overhead that has caused it to look bloated compared to the competition.

And AT&T; said it will no longer manufacture personal computers, opting to purchase the machines from others and resell them under its own brand. Analysts applauded the move, saying that even a more streamlined AT&T; had little chance in a business where profit margins can be as low as single digits and where new product cycles occur every six months.

Instead, AT&T; will concentrate on markets where it has enjoyed a following: Computer systems customized for banks and retail stores, the customers that are the legacy of NCR Corp., the Dayton, Ohio, company that AT&T; acquired four years ago after making a hostile bid.

Just how quickly the spinoff can be completed will hinge in part on how successful these resuscitation efforts prove to be. “We want them to be financially strong when we launch them,” Allen said of the new companies to be created by the break-up. Both the telecommunication services and equipment businesses are robust.

Given that AT&T; at one time had ambitions to be the premier computer company in the world--bigger even than IBM--its current goals are modest indeed.

*

AT&T; had once considered computers a natural extension of its telecommunications business. And that’s not as illogical as it might now seem: Consider that the central switch--the box that routes telephone calls--is in essence a highly specialized computer, and it is easy to understand why AT&T; expected to do well in the computer business.

Indeed, AT&T; was responsible for some true breakthroughs in computer technology. Bell Laboratories, its research arm, had invented the transistor and the Unix operating system, software that is popular with engineers and scientists. For decades it produced a broad range of computers for internal use.

“They thought they were well-positioned to be the premier computer company, a real challenger to IBM,” says John Logan, a one-time AT&T; consultant who is now an executive vice president with the Aberdeen Group, a Boston market research firm.

A consent decree signed with the U.S. Justice Department in 1956 prohibiting the phone giant from marketing computers outside its own company was an obstacle to its grandiose plans. But that was removed in 1984 when a federal judge ordered AT&T; to cut loose its 22 regional Bell operating companies. That same ruling said AT&T; would no longer be barred from the data communications business.

AT&T; wasted no time introducing its first computers, a line of mid-ranged machines that ran its Unix operating system. Unfortunately for AT&T;, its debut in the computer market coincided with an industry slump. But AT&T; had other problems, among them that its factories were unable to make machines as quickly--and consequently, as cheaply--as the market demanded.

AT&T; began turning to other computer companies--like Italy’s Olivetti and Silicon Valley start-ups Convergent Technologies and Sun Microsystems--for help. (In return, AT&T; often made a multimillion-dollar investment in the manufacturer.)

Only two years after its first machines were introduced, AT&T;’s computer business was in deep trouble, losing about $1 billion a year. Layoffs and other cost-cutting measures were undertaken, but the company remained without a coherent strategy.

AT&T; thought it had found one when it acquired NCR in 1990. AT&T; had been impressed with the plan that NCR had laid out for its computer business: New machines would be based on “open systems,” a term used for computers built with widely available hardware--like Intel Corp.’s microprocessors and software like AT&T;’s Unix.

*

After several months, AT&T;’s hostile bid for NCR was finally completed at a cost of $7.4 billion. But if the phone giant thought that buying NCR meant the end of its problems, it soon realized it was mistaken.

NCR chafed under AT&T;, resulting in the loss of key managers. And AT&T; tried to transform NCR into a computer company with broad appeal. The end result was that AT&T; spent millions with little to show for it and NCR’s longtime customers found themselves neglected.

Ironically, it is those customers that AT&T; is hoping will remain faithful during the spinoff.

“The best thing senior executives can say right now is, ‘We don’t know,’ ” said Aberdeen’s Logan. “It’s going to be a 10-month process. New accounts aren’t going to make a multiyear commitment until everything becomes much clearer.

“The best they can hope for is that the longtime accounts stay on board.”