Failure Is an Option

There were problems with the clones too. Power Computing and Umax both said that, according to their licenses with Apple, they couldn't use Exponential's chips unless Apple specifically agreed to let them do so -- even though they had already placed purchase orders. Moreover, Apple would have to agree to reengineer a small but significant portion of its machine's architecture, called ROM, thereby allowing the clone makers to alter their Mac-compatible motherboards. Only then could the clones fit the Exponential chip into a "common reference hardware platform." "You couldn't just drop this thing in,'' explains Jeff Thomas, 43, who served as vice president of engineering at Exponential.

On March 3, 1997, Campbell and Shriner again met with Rubenstein, who reiterated what the whole world knew by now. Apple was in a serious financial predicament. Drastic measures had to be taken, among them a reduction in Apple's engineering staff. Then Campbell and Shriner received a message that they would not have believed just a month earlier: Apple could no longer sustain an engineering team for the purpose of enabling Exponential chips to be built into Macintosh systems. Despite its purchase agreements, its financial backing, and its seat on Exponential's board, Apple was signaling that it would not be a customer for Exponential's product.

It was a crushing blow -- but for Apple, it might have been the only option. "Clearly, Apple had its back to the wall," says Tim Bajarin, one of Silicon Valley's most respected analysts. "It was in survival mode. The clones were eating its lunch, and so were the Intel guys. On the other hand, from a design and performance model, the Exponential chip made a lot of sense. Anyone who comes to you and says, 'We've got a 400- to 500-MHz chip,' gets your attention.'' Shriner and half a dozen members of his team -- including Dorris, Thomas, Valdes, and Peter Mehring, who was then VP of research and development at Umax -- insist that a commercially viable chip was within reach. "I felt like I'd been knocked off my horse,'' Shriner says. "We'd solved our problems. We were nearing completion of a second-generation chip that would have attained our goals. We were doing what everyone had agreed that we would do. And then they killed the deal.''

The Reckoning

"I think this is going down."

Shriner returned to San Jose. he stood for a moment in the parking lot, composed himself, and went into Dorris's office to ask his CFO how long the company could keep going. "Our feeling was that if we could just get the chip out the door, we'd be able to rescue the company," Shriner says. Dorris had already been on the phone with Campbell. Unless Exponential got a quick infusion of cash, she told Shriner, the company would issue its last paychecks on April 1, which was a little more than three weeks away.

Shriner called a staff meeting, assembling his top team in the company cafeteria while Dorris called Nixon in Austin. "We thought we'd better face this head-on and inform everyone about where we were going,'' Shriner remembers. He conveyed the news of Apple's decision but reminded his staff that Exponential was still negotiating with the clone makers. Besides, Shriner added, he and Dorris were about to set out on a third round of fund-raising. They hoped to bring in $15 million to $20 million.

Amazingly, the staff meeting didn't turn into an insurrection. It turned into a declaration of resolve to produce the world's fastest chip. "We were within weeks of laying people off,'' Shriner says. "But people on our team told us that they wanted to keep on pushing. I knew I had to do whatever was possible.''

So Shriner and Frank Marazita, 43, Exponential's vice president of manufacturing technology, went to Tokyo -- where, as Marazita puts it, "we fell on our swords.'' They told Hitachi executives about Apple's pullout and begged for a break on prices. "There were 12 executives on one side of the table," Marazita says. "They just got up and left us sitting there for two hours. When they came back, they told us they would lower the prices.''

Bolstered by that news, Shriner and Valdes went back to John Rubenstein and asked him to remove the obstacles that Apple had imposed on the clones when it had refused to reengineer the ROM. If Apple wouldn't do the engineering work, Shriner asked, would the company at least release the license to the ROM so that the clones -- if not Exponential itself -- could do the work on their own? The idea went nowhere. Apple's intention was to begin charging higher prices for the Mac operating system, not to let clones engineer around it. "We had been traveling 100 miles an hour,'' says Shriner, "and then we hit a wall. Without the ROM, we couldn't work with the clones. Without the clones, we were about dead.''