Last June, just after receiving an award for making the greatest contribution of any employee, Lu turned in his resignation at IBM. A month later he put together a seven-man team and returned to Taiwan to start Estron Inc., which aimed to develop a new memory chip with 16 million bits.

Not yet 40 and already research and development manager of IBM's technological product department, Lu left the state-of-the-art research facilities at IBM for the Ministry of Economic Affair's "Submicron Project."

A Gamble for Seven Families: While the domestic computer industry is developed, production of the key upstream component of memory chips is dominated by the United States, Japan and Korea. In order to break free from foreign control and establish technological independence, the Ministry of Economic Affairs, the Industrial Technology Research Institute and industry have combined forces under this plan, which is estimated to cost NT$7 billion. When the director of the Industrial Technology Research Institute, Chang Chung-mo, went to the United States looking for talent, he offered Lu the job of heading up the nation's drive toward developing upstream products.

"I was tempted by the idea of participating in the nation's development of an industry that is essential for international competitiveness."

Confident in his own abilities and that his partners, all of whom have an excellent grasp of technology, were the best possible choices at the time, Lu nevertheless had serious doubts about the political situation and was worried by rumors that certain companies were going to move overseas because of the faltering economy. "This industry is high investment and high risk. If government quavering makes things go sour, it means throwing billions of dollars into the sea." Hence, when he started, he really felt that he was "betting with the future of seven families."

Lu's ace in the hole was simply rolling up his sleeves and working with over 20 young engineers seven days a week, fifteen hours a day. "With such intense competition, there's a new generation in electronic design every three years. Only by putting in more hours can we catch up," he says.

Winning the First Round: In June, he won the first round. In less than a year, Estron Inc. had developed a 256 K SRAM (static random access memory) device that was, at 12 nanoseconds, the world's fastest. This device is needed by the new generation of high-function microcomputers.

Back in Taiwan for more than a year, there is still one thing that Lu can't get used to: the traffic. Driving home late at night, he ioften given a fright when cars coming from the other direction illegally pull into his lane. Living near the Hsin-chu Science Park, far from the grime and noise of Taipei, he is nonetheless dismayed by the disorder of the Legislative Yuan, the lack of respect for the law, and the rampant speculation in the economy. "I'm really not willing to see Taiwan's power be slowly and meaninglessly sapped away," he says.