The chips are down - so look to the future

The chips market is now about much more than PCs fuelled by Intel and x86 clone processors, just as it was before the last 15 years. Here, Sonya Rabbitte casts her eye over a battered landscape populated by still upbeat companies including ARM, IBM, Intel and Sun...

Cautious investors breathed a sigh of relief this week as chip bellwether Intel posted a 96 per cent fall in profit - actually meeting expectations - and IBM managed to beat analysts' forecasts, sheltering the dismal performance of its microelectronics business among the more robust earnings of its services and software arms.

Chip stocks rallied on global exchanges as shareholders all shared the same thought - it could have been worse. A trough in the third quarter must mean things can only get better. Right?

Gartner Dataquest estimates semiconductor sales are down 30 per cent this year to about $168bn. That's the good news.

The bad news is that a 30 per cent downturn is an optimistic forecast, dependent on 10 per cent sequential revenue growth between the third and fourth quarters of this year. Flat sequential growth will see annual growth fall by 34 per cent.

In the middle of a chip downturn the 42 per cent jump in revenue posted by ARM this week seems like a miracle on a par with the loaves and fishes.

ARM - a chip design company - makes its money by developing new chip designs and licensing its technology to several dozen manufacturers.

It's no accident that while 58 per cent of its third quarter revenue came from licence fees, revenue from royalty fees - the money it receives when a chip based on its design is shipped - fell by 12 per cent.

ARM's success is a clue to the future direction of the chip industry, says Alan Brown, an analyst with Gartner Dataquest.

He believes ARM defied the trend not because of its business strategy but because of its size.

"In a recessionary market there are likely to be the big players and then some small niche players, but the mid market won't survive," he said.

The sort of companies Brown envisages disappearing are those with revenues in the couple of hundred million dollars range while more of the larger players will follow the example of Motorola and Siemens, commercialising their internal chip divisions.

The PC processor will take a back seat in the post-recessionary chip market as both big and little players refocus on emerging technologies.

Research and development teams at IBM and Intel are already working on chips for wireless infrastructure, networking and flash memory. IBM is hedging its bets on a new chip material, silicon germanium, which will allow high-end chips to process at much greater speeds than even advanced copper chips.

IBM has for the most part avoided the commodity chip market - chips for PCs and other consumer electronics - in favour of producing high-end components.

Rubert Deighton, communications manager for IBM Microelectronics, believes this focus will pay off in the next 12 to 24 months as demand for networking and wireless chips grows.

"Telecoms infrastructure and network interconnection is critical. We want to link up with the internet on our mobiles and hand-held devices but the infrastructure isn't up to where it should be. That's why you see telecoms in chaos and Cisco up the wall. There will be tremendous growth in this area," he predicted.

Intel also sees itself competing in this market. Short-term growth hinges on increased sales of Pentium 4 but the medium- to long-term growth will come from the telecoms and wireless sectors.

"The current environment has pushed out expectations of an immediate upturn, and broadening our core business and growing in other areas is part of our strategy," said John Woodget, Intel director of sales for Northern Europe.

He added: "The communications market is not providing any return on investment today but we expect to reap the benefits in an upturn."

Rival player Sun is also preparing contingency plans for an upturn. Sue Kunz, global marketing manager for the company's processor product group, says the future demands a day-by-day approach, but holds many possibilities.

For Sun there is the 64-bit battle with Intel, and the aim to out-do Intel's 2GHz McKinley processor.

"We've made strategic decisions to make sure that when the market up turns we're there with all the right solutions. We're prioritising investments and keeping intellectual property - we're not the sort of company to go radical and ditch infrastructure that has worked for years," Kunz said, referring to Compaq's recent decision to ditch its own Alpha chip.

Across the board, the plans for medium-term growth seem solidly in place but the short term is less sure. A lot of crystal ball gazing is offered in place of firm predictions for the next 12 months.

Nobody wants to be drawn on upturn forecasts when the chips are very obviously down. Everybody knows that - even for a rich company like Intel - a 96 per cent drop in profits is not a good market reflection regardless of how many billions are left in company coffers.

Gartner is brave enough to predict the industry has reached rock bottom and that a turning point is due in March 2002, although Brown warns it will be single digit growth rather than the 15 to 20 per cent witnessed in last years chips boom.

Whatever the next phase of growth, it will certainly be different. IBM's Deighton summed up the consensus, saying: "It's clear that the PC market is no longer the epicentre of the chips world. You don't need to buy a PC if you want to link up to the internet."

It seems we may be getting beyond the PC, if not beyond the processor.