Tuesday, January 20, 2009

Surviving the Crash

As of this writing, the papers are celebrating the heroic U.S. Airways pilot who successfully crash landed his plane in the Hudson River, saving all 155 passengers on board. Will the captains of the electronics industry be able to save their own ships from the current economic crash?

While the electronics industry in general and the semiconductor industry in particular have had recurrent ups and downs, there is little doubt that this down cycle is going to be one of the worst ever. Both industries are likely to look very different when they emerge from the trough.

Forget Startups

According to the National Venture Capital Association, VC fund-raising fell 71% in the fourth quarter of 2008 due to the recession and aversion to risky investments. For all of 2008, U.S. VCs pulled in just under $28 billion, down 21.4% from the $35.5 billion raised in the prior year. This money is largely going to established startups who are already selling a product and who, with some more marketing oomph, are likely to make a profit. On top of this, in all of 2008 there was only one initial public offering (IPO) to come out of Silicon Valley. With investors risk averse and no exit strategy in sight, VCs are becoming risk averse, too. New startups are toast. This will starve the industry of both innovation and the acquisition targets that enable larger players to innovate fast enough.

Forget Profits

This almost goes without saying. Innovation is the key to profits, since consumers are willing to pay a premium for cool new features. But competition always drives down price; if you can’t continue to compete on features, you’ll soon find yourself in the death spiral of competing on price. During a recession, it’s almost impossible to avoid the latter. It becomes a matter of survival of the fittest, and major industry consolidation is already underway. In good times, the market will penalize you for sitting on a pile of cash; now only companies with large cash reserves and well diversified cash flows are likely to survive. One financial analyst summed up the situation by describing his recommended positions during this recession: “Cash and fetal.”

Go Fabless

The days of “real men have fabs” are over. The CAPEX is unsustainable and TSMC can do it better than you, anyway. AMD has pushed its fab off its balance sheet, talking Abu Dhabi into investing in it (good luck, guys). Cypress is outsourcing all fab work under 90 nm, and TI is doing the same at the 45 nm node. Except for hanging onto an R&D fab—Cypress has even managed to monetize theirs as the Silicon Valley Technology Center—there’s no excuse for not diverting the CAPEX you’d otherwise spend on a fab into R&D. This is a game only the big dogs can play, and even they’re banding together to spread the pain around.

Get out of DRAM

If there was ever a commodity product, it’s DRAM. Everyone involved with it is losing money; Qimonda is in such dire straits that its owners can scarcely give it away. Micron converted much of its DRAM fab capacity to manufacturing flash memory, which until recently was a license to print money; now flash looks to be the next DRAM.

In the 90’s Korean companies undercut the Japanese to take over the DRAM market from them; now the Taiwanese are out to do the same to the Koreans. They’d better look over their shoulders at the Chinese. It’s time to end this losing game and get out of DRAM. Let SMIC or Grace make it all. Good riddance.


To achieve the critical mass to survive an extended downturn, consider merging with your competition. Lattice and Actel both have similar technologies, and both continue to watch Altera and Xilinx walk away from them; they’d be stronger if they combined forces, just as Lattice previously did with Vantis.

Cypress and Micron would be another logical pairing, though it’s difficult if not impossible to imagine T.J. Rogers and Steve Appleton jointly running the company. When you factor in executive egos, you’re more likely to see consolidation take the form of acquisitions, resulting in the industry having fewer if larger players after it all shakes out.

I’m beginning to understand why the ancient Chinese saying, “May you live in interesting times,” isn’t a blessing but a curse. These are going to be interesting times. Fortunately, like everything else, they’ll pass.

Friday, January 9, 2009

CES Thursday

I arrived at the South Hall to pick up my press pass with the usual fear and loathing: I don't gamble and I hate crowds, both of which define Vegas. Fortunately the crowds this year at CES are a lot smaller than last year, and the number of startups willing to gamble their annual marketing budget on a booth has also diminshed. There's still plenty worth seeing, but at least you don't feel quite so much like a salmon swimming upstream when you go to see it. I'm interested to see whether large trade shows like CES can weather the coming economic storm or instead blink out like once-mighty Comdex. My guess is that small local shows are the wave of the future and large shows the new dinosaurs.

My vote for the most fun exhibit is Nvidia's 3D demos of Guitar Hero. People were standing in line to grab a pair of 3D glasses to watch a computer-generated version of Sting rip through some hot licks. OK, so these guys are just chipheads, but they definitely know something about demand creation.

Qualcomm's demo of Android running on their Snapdragon chipset was also noteworthy, if not being quite as snappy. T-Mobile's G1 handset, made by HTC, may be the first to run Android--using Qualcomm's chipset--but Motorola, Samsung and numerous other manufacters have announced pending products based on the Android platform. Expect Android to unleash a wave of new handset applications, drive pricepoints down and cause heartburn for carriers whose 'walled gardens' will be under attack.

Steve Ballmer's opening keynote painted an optimistic future for technology: "No matter what happens with the economy, our digital lives will only get richer." The big opportunities for the tech industry are the convergence of the PC, phone and TV; a more natural interaction with devices that will incorporate speech and hand gestures; and 'the connected experience between devices' ("Was it good for you, too?"). Windows 7 beta will be available shortly and will "deliver the ultimate Windows experience." One can only hope that will be resemble Windows embedded and not Vista.