SANTA CLARA, Calif.-- While posting record fourth-quarter results today, Intel Corp. announced plans to increase capital spending 47% to $5.0 billion in 2000 from $3.4 billion in 1999. The world's largest chip maker also said it will spend $3.8 billion on R&D this year, an increase of 22.6% over $3.1 billion in 1999.
Intel is planning to make those big investments after recording record revenues of $8.2 billion in the fourth quarter and beating Wall Street's estimates for earnings in the period, ended Dec. 25. Intel's fourth-quarter revenues grew 8% from $7.6 billion in the same period a year ago. Including acquisition-related costs, Intel's net income was up 2% to $2.12 billion compared to $2.06 billion in the 1998 period.
"In 2000, we look forward to continued growth in our core microprocessor business, and to the mid-year production of our new Itanium processor which began sampling in the fourth quarter of last year," said Craig R. Barrett, president and CEO of Intel. "We are also excited about the expansion of our new Internet-related businesses with a series of acquisitions in 1999. This year we expect to grow revenues in our networking, communication and wireless businesses by 50% or more," he added.
Intel's $5.0 billion budget for capital expansion is nearly twice the size of any other semiconductor company, according to estimates by Dataquest. The San Jose market research firm believes Samsung Electronics, Taiwan Semiconductor Manufacturing Co. (TSMC), and United Microelectronics Corp. (UMC) will be in a virtual tie for second place with capital budgets of about $2.5 billion in 2000.
The huge increase in capital spending by Intel will be used to deploy additional 0.18-micron process capacity as well as begin the move to next-generation 0.13-micron technologies and larger 300-mm wafers.
"If we need more 200-mm capacity we will add it, and we will move rapidly to 300-mm at the 0.13-micron generation," said Michael R. Splinter, senior vice president and general manager of Intel's Technology and Manufacturing Group, while speaking an industry conference earlier this week.
Intel is now beginning to see cost advantages in using 300-mm diameter wafers as device feature sizes shrink to 0.13 micron, said Splinter during his key note presentation at the Industry Strategy Symposium in Pebble Beach, Calif., on Tuesday. Splinter said Intel had been able to save money by reusing and upgrading 200-mm fabs during the late '90s, while minimum feature sizes shrunk from 0.25 to 0.18 micron. But costs factors are changing as new semiconductor materials--such as copper and new dielectrics--and process steps are needed to push feature sizes to the next generation, Splinter told capital equipment suppliers at the ISS meeting.
Wafer diameters of 300-mm "will essentially move back the capital cost per die to the early days of 200-mm," Splinter predicted. Currently, analysts disagree on when 300-mm wafer processing will take off, but most believe the economics will begin to shift sometime in the next three to four years (see today's story from the ISS conference).
Intel plans to first launch 0.13-micron technology on 200-mm wafers and then transition production to 300-mm substrates before that process generation reaches peak volumes, Splinter said. Last June, Intel announced plans to start 300-mm production on a 0.13 micron process with copper interconnect in 2002, about one year after it begins 0.13-micron production on 200-mm wafers (see June 9, 1999, story).
