The Rollercoaster Ride of UMC, Taiwan’s Original Chip Champion
While TSMC dominates today’s semiconductor industry headlines, its lesser-known sibling United Microelectronics Corporation (UMC) wrote the first chapter of Taiwan’s chip saga. Founded in 1980 as a public-private hybrid — backed by Taiwan’s National Development Fund, Dutch giant Philips, and local banks — UMC pioneered the island’s shift from government labs to commercial chip production. But its early years read more like a corporate rollercoaster than a triumph of technocratic planning.
When Dr. Robert Tsao stepped in as general manager at the end of 1981, he found himself at the helm of a ship with sails but no wind. UMC was an Integrated Device Manufacturer (IDM), much like Intel or AMD, juggling the roles of designer and producer of silicon chips. However, the company’s initial product lineup of low-cost chips for digital gadgets like calculators, watches, and musical greeting cards was not exactly the stuff to keep its expensive fab running at full steam.
The early days were tough. With Taiwan’s economy in a slump and the global semiconductor market hitting a low point, UMC was hemorrhaging cash, closing 1982 with a net loss on meager sales.
The following year, UMC enjoyed a stroke of luck. The 1983 breakup of AT&T’s Bell System monopoly created chaos in the U.S. telecom market. Suddenly, every Baby Bell needed phone dialer chips — and fast. Established US and Japanese semiconductor firms were caught flat-footed and UMC’s factory went from idle to overdrive, pumping out 24 million chips (up from 4 million) and raking in $1 billion practically overnight.
Unfortunately for UMC, the party didn’t last. By September 1983, the demand for IC dialers crashed, the company’s fabrication facilities silent for almost half a year. By 1984, production volumes had plummeted back to where they started, leaving Dr. Tsao scratching his head over the future of UMC’s business model.
As UMC cycled between feast and famine, Tsao concluded that the company should shift to what he called an “OEM foundry” model. Instead of doing everything in-house, UMC would focus purely on manufacturing, while partnering with local design firms. This, he argued, would allow UMC to use its resources more effectively, avoiding the boom-and-bust cycle of relying on a single hit product to drive demand.
Although Tsao’s proposal was initially welcomed by the Ministry of Economic Affairs, it hit a brick wall when he was asked by the government’s technology tsar K.T. Li. to present it to Dr Morris Chang, who had just been appointed President and COO of General Instrument in the U.S. After Chang described the proposal as unworkable, ministry officials abandoned the proposal, sparking bitter accusations from Tsao that Chang stole his idea when he went on to adopt his “pureplay foundry” model at TSMC in 1987.
For his part, Dr Chang has always maintained that he conceived pureplay foundry concept while he was working at Texas Instruments long before Dr Tsao proposed his “OEM foundry” model. Although the two men have reconciled in recent years, their early rivalry underscores the fierce competitiveness that transformed Taiwan into a semiconductor titan, with TSMC and the much-smaller UMC now collectively commanding over 60% of the global foundry market.
