While hundreds of thousands of U.S. troops deployed to Saudi Arabia in preparation for war, Frohman was distracted by the risk Intel was undertaking. That gamble was a product of IBM’s decision, in 1980, to give Intel its big break, choosing the 8088 chip to power the IBM PC. But the computer giant had forced Intel to license its technology to a dozen manufacturers; even though Intel had designed the 8088, IBM thought it was risky to rely on Intel alone to manufacture the chip. So Intel was able to earn only 30 percent of the total revenues. Security and price leverage for IBM meant lower profits for Intel.
In 1983, with the 286, its next-generation chip, Intel had managed to convince IBM to cut the number of manufacturers to four, thereby increasing Intel’s own share of the work. And by 1985, after investing $200 million and four years of development in its even faster 386 chips, Intel had been prepared for a gamble. This time, IBM had acquiesced to Intel’s request to become the sole manufacturer of the chip that would power most of the world’s new desktops. This strategy would maximize Intel’s profits, but also its risk. What if Intel could not ramp up its manufacturing capability in time? And the bigger risk was the decision made by Intel’s management in Santa Clara to center much of this new responsibility in Israel.
The main burden fell on Intel’s Israeli chip plant in Jerusalem, which produced about three-quarters of Intel’s global output by running two twelve-hour shifts, seven days a week.
But now that output was under threat. Saddam Hussein had declared that if the United States launched an offensive, he would respond with missile strikes against Israel.
The Israeli government took Saddam at his word. Iraq had Scud missiles that could reach Tel Aviv in under ten minutes, and those missiles might be armed with chemical warheads. In October 1990, the Israeli government ordered the largest distribution of gas masks anywhere since World War II.
It was a surreal time in Israel. In kindergartens, teachers showed five-year-olds how to put on their gas masks in case of attack, and everyone practiced rushing to specially prepared “sealed rooms” if the sirens went off. The distribution system for the masks was elaborate, with every household receiving a note in the mail telling them where they could pick up the equipment. The IDF placed its Home Front Command offices in malls, so it was not uncommon to pick up some new shoes and a cup of coffee along with a set of gas masks for the whole family.
Frohman did what every Israeli manager does during or in advance of war: he drew up contingency plans for the “standard” war scenario, in which employees would be called up for reserve duty. Most Israeli men under forty-five serve in the reserves for one month every year. During an extended war, these civilian-soldiers can be called up for as long as the government deems necessary. This exacts a huge economic toll on businesses in Israel—including lost work days and less productivity—even during peaceful times. During a war, employees can be absent for weeks or even months. As a result, some Israeli businesses go bankrupt during war.
In early January 1991, U.S. and European commercial airlines suspended or curtailed their flights to the region. On January 11, four days before the United Nation’s deadline for Iraq to withdraw from Kuwait, the U.S. government advised its nationals to leave Israel. On January 16, the Israeli government announced that all schools and businesses, except for certain essential enterprises (the electric utility, for example), must close for the week and maybe longer. The government wanted people at home, off the roads, and poised to hop into their sealed rooms at the sound of air-raid sirens.
For Frohman, compliance with the government’s directive would mean suspending the production of Intel’s 386 microchip at a critical moment for the company. Frohman expected to have management’s full support for a shutdown, but he also knew that just because an employer is willing to grant an employee sick leave, it does not mean that their relationship will go on unaffected. Especially when the “ailment” is one that could conceivably repeat itself in the future.
“We already had a number of struggles inside the company over the transfer of strategic technologies and critical products to the Israeli operation,” recalled Frohman. “I was convinced that if we had to interrupt production, even for a brief period of time, we would pay a serious price over the long term.” Frohman had expended time and political capital to persuade Intel’s management to put the future of the company in the hands of an overseas outpost, a dream of his since he’d first left Intel. And it was this outpost that was about to find itself on the receiving end of Scud missiles.
But Frohman had another—surprisingly far greater—concern: “I kept thinking about the survival of Israel’s . . . still small high-tech economy.” The key stumbling block to further investment in Israel was the lingering impression of geopolitical instability in the region. If Intel couldn’t operate in an emergency situation, then any confidence that multinationals, investors, or the markets had in Israel’s stability would instantly crumble.
Frohman had spent enough time abroad to be familiar with the rap against investing in Israel. Almost every day a bad headline about Israel ricocheted around the world: another terrorist attack . . . another provocation on its border . . . more bloodshed. Intifada. Violence, terror, war. It was the only narrative people knew.
He believed that both Israel and its economy needed a counternarrative. As the January 15 deadline approached, he became fixated on an imaginary boardroom debate—taking place somewhere in the United States—between an executive who was enthusiastic about investing in Israel and a cautious board that thought he was reckless. What would the enthusiast need in his back pocket? I understand your skepticism. I saw the news, too. But let’s not forget that Intel was producing the 386 chip—one of Intel’s most important microchips—in Israel during the Gulf War, and the Israelis never missed a beat. They stayed on schedule. They were not late . . . not even once . . . not even when missiles were falling on them.
On January 17, Frohman informed his employees of his unilateral decision to keep Intel Israel open during the war, in defiance of government orders, but on a voluntary basis: no worker would be punished for not showing up.
At 2:00 a.m. on January 18, Frohman, like most Israelis, was awoken by air-raid sirens. He and his family quickly put on their gas masks and sealed themselves into their home’s safe room. When the all clear sounded, they learned that eight missiles had struck Tel Aviv and Haifa—near Intel’s main R&D facility—but they had not been armed with chemical warheads. More missiles were expected in the days ahead. Whether Saddam would arm future Scuds with chemical capabilities was still unclear.
At 3:30 a.m., when Frohman arrived at the plant with his gas mask, he went straight to the clean room—the heart of the chip factory, where, to maintain a dust-free environment, technicians worked in sealed suits that made them look like astronauts. Work there had already resumed. He was told that when the sirens had sounded earlier, the employees had gone to a sealed room in the plant, but after quick calls home, they had returned to their work stations. When the first postattack morning shift began, Frohman expected to see—best-case scenario—half of the shift; 75 percent showed up. Following a second Iraqi missile attack the next night, turnout at Intel’s Haifa design center increased to 80 percent. The more brazen the attacks, the larger the turnout. Welcome to Israel’s “new normal.”