Posted on 09/22/2002 12:33:36 PM PDT by j271
SANTA CLARA, Calif. — The bear may be heading back to hibernation. The electronics industry is beginning to shake off its bear-market blues with a pair of assessments that suggested resurgent growth next year — two years after the industry's worst recession began.
Semico Research Corp. (Phoenix), at its 2002 Forecast Workshop here this past week, repeated its call for 30 percent growth for semiconductors in 2003. And LSI Logic Corp. chairman Wilf Corrigan, in a presentation to financial analysts at the New York Stock Exchange, said the market should grow 20 percent next year.
Those optimistic views came amid the continued decline of tech stocks in the equities markets and head-scratching short-term forecasts from National Semiconductor Corp. and Intel Corp.
"The uncertainty of the PC market, combined with the mostly crappy worldwide economy, leaves us cautious," National Semiconductor chief executive officer Brian Halla said.
And Intel dropped its third-quarter revenue projections to the low end of expectations. The move came a week after CEO Craig Barrett, while on a tour of Asia, said, in effect, that it was anyone's guess when IT spending would resume.
But many in the industry are looking beyond the carnage and hanging their hats, albeit gingerly, on next year.
|
| Semico sees a dramatic end to the pain next year and in 2004. The next dip comes in 2005. |
Semico Research lowered its overall semiconductor forecast for 2002 but predicted that the market will rebound and recover in 2003 and 2004. In May, Semico had forecast that chip sales would bounce back from the severe 2001 recession with 17 percent growth in 2002. Earlier this week, the market research firm lowered its overall forecast for this year, saying that the worldwide semiconductor market will grow only 6 percent.
The market fell in 2001 by a staggering 31 percent from 2000 figures, according to Semico.
But the firm sees 30 percent growth in 2003 and a strong market in 2004. The next downturn will arrive in 2005, said Semico president Jim Feldhan.
Feldhan also put a sobering spin on the outlook, saying the industry will have to wait two more years to see worldwide semiconductor sales return to the peak levels achieved in 2000. "It won't be until 2004 before we get back to [where] we were in 2000," he said at the conference. "Then, we're back in a downturn in 2005."
In an interview, Feldhan also expressed some concerns about the worldwide IC market in 2002. "Prices did not firm up like we thought," he said.
On the other hand, IC unit growth is expected to hit a healthy 11 percent in 2002, he said. And the 2003 market will reflect the events of 2002.
"We will see a flat first quarter [in 2003], then things will accelerate in the second half of next year," he said in his keynote address.
What will drive the market in 2003? While the PC and cellular-phone upgrade markets will propel demand, China will also remain a bright spot. "If we look into the next decade, China will become the next golden economy," he added.
Others agreed — at least to some extent. IC growth in 2003 will depend upon several factors, said Ron Leckie, an analyst for market research firm Infrastructure Inc. "It really depends upon pricing [for ICs]. If corporate spending comes back, then we could see 30 percent in 2003."
Dave Chavoustie, vice president of sales for stepper maker ASML (Tempe, Ariz.), sits in the camp that seems both amazed and befuddled by the recession. Chavoustie, who has worked both on the chip front (via stints at Advanced Micro Devices and VLSI Technology) and in semiconductor gear, said the reports from his field force are confusing at best.
"We have some customers who were about to pull the plug on an order suddenly call and say, 'Hey, we need these machines right away.' We have others who were about to take shipments [tell us to] hold off. Then they call a few weeks later and say, 'Now we need them.' It's crazy."
Where the action is
Semico's lofty predictions for 2003 see the computer segments accounting for some 53 percent of worldwide IC sales. Communications will claim about 18 percent of sales, consumer gear 13 percent, industrial equipment 8 percent, automotive 6 percent and military apps 1 percent, the researchers said.
Among the bright spots in the industry are DRAMs, which will grow by 20 percent next year. Semico puts the DRAM category's sales numbers at $11 billion in 2001, $16 billion in 2002 and $20 billion in 2003. Longer term, it expects another growth year in 2004, with DRAMs logging $24 billion; a $22 billion DRAM market in 2005; and a rebound, to $27 billion, in 2006. The key growth driver is the shift to double-data-rate (DDR) SDRAM, especially at the 256- to 512-Mbit densities.
Semico's Joanne Itow last month said she was sticking to her foundry-market forecast for 2003. She predicts the foundry business will grow by 34 percent next year over 2002.
Leading-edge capacity remains tight right now, Feldhan said. "The end result is that foundry capacity will be in short supply," he said. "Leading-edge processes will be tight in the second half of 2002 and 2003."
Corrigan, in his presentation to analysts, was cautiously optimistic that the worst of the slump is over, that most of the extra inventory is out of the pipeline and that LSI Logic will meet third-quarter 2002 guidance calling for sequential revenue growth of about 10 percent. "We are on track to return to pro forma profitability in the fourth quarter," he said.
It took a fair amount of doing. In the past two years, the company has redistributed its market portfolio among consumer, storage and communications to de-emphasize communications from a 50 percent share of the pie in 2000 to about 19 percent this year. Commensurate with the shift, the less volatile consumer and storage markets have grown to account for 31 percent and 51 percent of LSI Logic's business, respectively. In addition, the company cut its staff by 50 percent, to 5,300 employees worldwide, among which there are still a healthy number of engineers — about 3,000.
"I want to emphasize this phenomenon: We need engineers to keep innovating with new concepts such as our just-released RapidChip silicon and methodology technology," said Corrigan. "RapidChip aims to fulfill the opportunity presented by a potential $11 billion market by 2006 that fits between high-performance ASICs at the high end and FPGAs at the low end of design complexity."
LSI Logic management hopes that RapidChip will add substantially to revenue in coming quarters. Chief financial officer Bryon Look reported to analysts that revenue in the third quarter has grown 10 percent and that the company has now experienced growth in four consecutive quarters.
"That, together with the 20 percent growth over the same quarter last year, bodes well for LSI in this stagnant economy," said Look.
He reported that while fab utilization fell to as low as 30 percent in 2000-2001, it is now at a healthy 75 percent. To maintain steady growth, LSI Logic has reduced capital-equipment expenditures to $25 million this quarter and will spend less than $100 million on capex all year. "We will be increasing significantly the use of outside foundries while limiting our operations costs," Look said. "Currently we are outsourcing 30 percent of our work, and we expect to increase that to 50 percent in 2003."
In the immediate term, Corrigan is hanging his hopes on the sale of DVD players and videogame consoles — two product areas where LSI Logic ASICs are dominant. "The DVD is a $50 million market and growing. This year both DVDs and videogames are going to make more money than feature films at the box office," Corrigan said. "For both, consumers need the players and consoles to play the DVDs and games. We expect a lot to be bought between Thanksgiving and Christmas."
— Additional reporting by Brian Fuller. Mark LaPedus is editor of Semiconductor Business News, an EE Times Network Web site.
You and Benjamin Braddock.....
semiconductors are becoming more powerful, more intelligent and smaller. track your child. monitor medical data in the body. send signals to the hospital when conditions worsen. sense radioactive materials in shipping. improve memory and reduce the price of computers.
just a few things a soybean can't do.
there will be a comeback, eventually.
No, thank you. You're right. Soybeans are much nicer.
Didn't say there wouldn't be a demand for them...there is usually demand for a 'commodity.' Only said that I wouldn't look to them to drive the economy, like they did in the bubble.
EXACTLY. Good post.
I hear 1950's style toasters are making a comeback too.
Especially when prepared as "edamame". Great appetizer!
What's needed is that good old "Yankee Ingenuity" to bring the manufacturing costs. The secondary benefit, (actually primary) would be the reduction in energy production needs. Since about 25% of energy use in the US is for lighting, this could be a place where significant inroads could be made in saving energy. Remember when Nader said the first step in any energy plan was to "change the light bulbs" well he wasn't kidding.
I worked in this industry until Feb when I was "census managed", (latest term for downsizing!). Unfortunately, their strategic planning is for sh&t since they are a commodity business and can't see beyond the next quarter's plan/results!
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.