Fujitsu Restarts Oregon Expansion
Engines with $550 Million Fab
Ah, the killer app: It's one of the most enduring phrases spawned by the Internet Age, and it's spurred many a blockbuster deal - not to mention the winning and losing of untold millions upon millions of dollars.
And the killer app du jour -- personal communications services (PCS) -- is just what's behind Fujitsu Limited (www.fujitsu.co.jp) and Fujitsu Microelectronics' (FMI at www.fujitsumicro.com) announcement of a US$550 million expansion of FMI's flash-memory chip operation in Gresham, Ore., located east of Portland. Ironically, Gresham is the very same site where an earlier killer app erased Fujitsu's expansion plans from the drawing board in 1997.
What's spurring Fujitsu to add 300 new Gresham workers is the resolutely bullish demand for the kind of flash-memory chips that are used in the increasingly ubiquitous array of PCS devices (e.g., cell phones, palmtops, etc.)
"Gresham will be one of Fujitsu's major flash-memory facilities and one of the most significant such facilities in the United States," says Kazunari Shirai, group president of Fujitsu's Electronic Devices Group in Japan. "Along with our FASL factories in Japan, we expect that Gresham will play an important role in Fujitsu's worldwide flash-memory operation."
Surging Flash-Memory Demand Puts
Fujitsu is only one of the many chipmakers that are betting their business fortunes on boosting production of flash-memory chips for the PCS market's juggernaut.
In fact, Oregon is experiencing a sort of fab-expansion fever. In addition to the major Fujitsu expansion in Gresham, Intel has acquired 140 additional acres (56 ha.) near its fab site in Hillsboro, Ore. And Hyundai MicroElectronics is seeking government approval to add as many as two additional fabs at its site in Eugene.
As those massive expansion plans indicate, market demand for flash-memory chips is surging. And that's supremely welcome news to a global semiconductor industry badly staggered by a worldwide sales slump during recent years.
That's also a hugely encouraging economic development for Oregon, as the chip industry accounts for a third of the state's high-tech jobs.
Demand for flash-memory chips has been so full-bodied, in fact, that Fujitsu tripled its initial expansion plans in Oregon. Announced four months earlier in January, the original expansion design called for boosting chip output in Gresham to 10,000 wafers a month. Explains Shirai, "The company decided to triple the goal to 30,000 wafers per month to meet the robust demand for flash memory in cellular phones, handheld wireless devices and other mobile products."
But while production capacity will triple, the new plan involves a more exponential increase in capital expenditures. Fujitsu's original expansion plan to boost Gresham's fab output to 10,000 wafers a month only called for a $60 million investment.
Fujitsu officials say that part of the expansion will entail equipping a vacant Gresham plant that had been mothballed during a business downturn. The expansion will also involve adding manufacturing equipment to a second existing plant, they add.
The expanded production will be conducted under a wafer-foundry arrangement with Fujitsu AMD Semiconductor Ltd., Fujitsu officials say. Fujitsu AMD Semiconductor is a joint-venture company created in 1993 between Fujitsu and Sunnyvale, Calif.-based Advanced Micro Devices.
Fujitsu's Gresham fab expansion also validates an initially uncomfortable change in corporate strategy.
Gresham, in fact, is a landmark in both the old and new strategies. Fujitsu opened a DRAM production fab in the Oregon city in 1988. Business blossomed over the next 10 years, and the initial fab operation in Gresham doubled in size to span 545,000 sq. ft. (49,050 sq. m.)
Then came a wicked shift in the market. In 1997, Fujitsu was poised for yet another expansion at the Gresham site. The stunning swing in the marketplace, though, forced the company to pull the plug on its expansion plans. Fujitsu bailed out of a property tax-break agreement it had reached with Gresham and Multnomah County. With the market's whiplash reversal of fortune, the company knew that it wasn't going to be able to create the new jobs to which it had committed. Business conditions, in fact, prompted Fujitsu to close its oldest Oregon plant in 1997.
The DRAM downturn also sent Fujitsu back to the drawing board. It came back with a new strategy: de-emphasizing the PC-centric main-memory DRAM sector and emphasizing flash-memory and higher-speed products.
The repositioning seems to be working. The Gresham expansion decision, in fact, comes on the heels of Fujitsu's return to profitability, which the company reported in late April.
By the old-line Japanese giant's standards, it wasn't a huge profit: $403 million on 1999 net sales of $49.6 billion, which matched 1998's sales levels. But Fujitsu's 1999 profit of $403 million must've looked pretty peachy compared to fiscal 1998's consolidated net loss of $128 million. And 1999's profit came despite what the company's official release terms "the negative impact [of] retirement benefit restructuring costs, as well as the influence of the high yen [and] a reduction in extraordinary losses."
As things have turned out, Fujitsu dramatically pulled back in the DRAM sector just as the roof was getting ready to fully disintegrate. 1999 worldwide DRAM sales totaled some $14 billion, according to IC Insights, a Scottsdale, Ariz.-based market research firm. That's a long, long slide from 1995's worldwide DRAM-demand peak of $40.8 billion.
Fujitsu officials attributed 1999's better business year to "some improvement in Japanese consumer spending on PCs, strong economies in the U.S. and Europe and healthier economic conditions in Asia." They added, though, that "growth in consolidated net sales was restrained due to such factors as the slow pace of overall economic recovery in Japan and Y2K-related constraints on IT investment worldwide."
The company's saving business grace, officials added, came from "strong U.S. demand for fiber-optic transmission systems and robust sales of flash-memory and logic ICs, [which] helped drive double-digit increases in sales of communications systems and electronic devices." Obviously, the Gresham expansion is designed to strengthen the flow of what's become a strong revenue stream.
Fujitsu Gets to Finish What
Fujitsu's big expansion in Gresham is also probably gives the company a psychological boost of a different sort.
Like most Japanese companies that have set up U.S. operations, Fujitsu has been a positive force in the communities in which it's located. In Gresham, it's been an active participant in a number of civic-minded endeavors.
Pulling out of the Oregon expansion in 1997, then, left the company dangling in what must have been a most uncomfortable position. The city and the county squabbled over the payoff on the incentives, despite Fujitsu's willingness to pay $30 million to $40 million in penalties. Now, though, the company is back in expansion mode in Gresham. And that expansion plan will begin in the Gresham facility that was closed during the seismic shift in DRAM demand. The 300 new jobs, which will be added over a three-year period, will increase the company's Gresham work force to 800 employees. Employees who are being let go by one of Fujitsu's sister companies, Fujitsu Computer Products of America, will fill some of the new positions. Fujitsu Computer Products of America is cutting 160 positions from its work force at its repair facility in Hillsboro, Ore.
Max Talbot, director of the Gresham Community and Economic Development Dept. (www.ci.gresham.or.us), says Fujitsu's expansion will mark a substantial addition to the city's employment base and property tax roll. The property tax increase alone could enable the city to increase services, he said.
Gresham Building Development Manager Rob Fussell calls Fujitsu's new positions "good jobs, the kind we're always happy to see come to town."
For one town, at least, the new killer app has brought a sweet, deep breath of economic life.
©2000 Conway Data, Inc. All rights reserved. Data is from many sources and is not warranted to be accurate or current.