Semicon West Takeaways
By David Lammers
Market uncertainty and equally squishy technology roadmaps hung over Semicon West 2011.
People left San Francisco questioning whether the formidable challenges related to the 14-11-9nm technology transitions will conflict with the move to 450mm wafers. Ditto for the shift to EUV lithography, which, many believe, has to take precedence over any wafer transition.
As West kicked off early last week, both Novellus and Applied held analyst events saying sales and bookings are taking a hit from low capacity-utilization rates at the major foundries (80 percent at TSMC, less than that at the GlobalFoundries fabs in Singapore) and weak DRAM sales due to single-digit PC growth.
“Our customers aren’t making any capacity buys right now,” said one manager at a major lithography vendor, whose discretionary spending budget is being slashed for the second half of the year. A manager at a major chemical company said his company has instigated a cost-cutting campaign for the rest of this year, while keeping watch on global economic conditions.
A manager at a large semiconductor equipment company agreed with the Novellus/Applied assessments, saying, “We are seeing some pushouts and are cutting back on costs. We are hunkering down right now.”
The hope is that this is a “pause” in chip equipment spending. But whether the pause becomes a downturn may depend on how the world financiers, politicians, and central bankers resolve the macroeconomic challenges.
There are plenty of reasons to be happy about the state of the semiconductor equipment and materials industries. Apple can’t make its iPad2s and iPhones fast enough.
In Austin, Samsung is planning to go ahead with its third major fab. While the Apple lawsuit over Samsung’s alleged phone/tablet copying is troublesome, that hasn’t changed Samsung’s collective mind about expanding in Austin, sources said. It makes perfect sense for Apple to swing some logic IC business towards TSMC and perhaps Intel as well. But the Apple pie (excuse the pun) is now big enough, and growing every day, that all of its vendors are among the fortunate.
And Intel’s recently updated plans to charge full speed ahead on its capex spending is good news for suppliers. That will lead to “a pretty darn good 2013,” one equipment company executive said.
Lithography for 450
There were some nice things said about the 450mm wafer transition at Semicon West. Mike Splinter said 450mm spending at Applied will exceed $100 million next year, for example, and others were conciliatory towards what has been an unwelcome subject thus far.
And there are concrete signs of progress. KLA Tencor introduced its 450mm-capable Surfscan SP3 unpatterned wafer inspection tool, for example.
The lithography stumbling block may see a temporary solution in the next few months. A contract with an unnamed lithography vendor for a 450mm-capable exposure tool is in the works, I was told. That would allow 450mm wafers to be patterned and then sent around to other equipment vendors for 450mm tool development. One ISMI source said an imprint lithography solution is the subject of the upcoming contract.
Also, Albany Nanotech is planning to build a 450mm development center that would support the ISMI 450mm development team. However, the Empire State is financially distressed, and politicians there are reluctant to commit to a 450-mm subsidy until the state figures out how to meet its current financial obligations. (Imec already has a 450mm-capable R&D fab in Leuven, and its program-based R&D model could accommodate a 450-mm tool development program, Imec managers said at Semicon West.)
Sematech’s member companies don’t want to pay for more than one-fourth of the 450mm equipment R&D spending. The SEMI members argue that the chip makers are the ones with the money to support most of the 450mm-related development costs.
If the financial story wasn’t complicating the 450mm agenda enough, there are the uncertainties about the technology and lithography transitions. One senior manager said the semiconductor industry simply cannot delay any technology transitions, since the gains from scaling are much more important than the cost benefits of using a larger wafer.
And since many companies are banking on EUV lithography at the 14nm node, the semiconductor industry will have to bring in 300mm EUV scanners before it seriously contemplates building full-scale 450mm fabs.
“EUV has to come before 450,” this senior manager said.
Not everyone agrees with that. Nikon and ASML could introduce 450mm 193nm immersion scanners fairly expeditiously, one source said. Also, lithography is not intimately tied to the wafer size, because it is a die-to-die process. ASML has built its EUV program so that the wafer handling robots and chambers could be upgraded to the larger wafers relatively quickly, another manager noted.
Finally, hanging over all of this is the question of what node will be reaching volumes when 450mm wafers are introduced. Let’s call this year, 2011, the year that 28nm technology ramps, leaving Intel aside for a moment. Then 2013 would be the ramp year for 22/20 technology, and 2015 the takeoff year for 14nm processes.
If you are a manager at Lam or Applied or TEL or Novellus or ASML or Nikon, do you bank on 2017 being the year that 450mm would be introduced for 11nm CMOS, with EUV a critical part of that process flow? Or do you hedge your bets and figure that the semiconductor industry will have its hands full bringing in EUV, finFETs/FD-SOI, TSVs, and other challenges, leading to a 450nm pushout to 9nm or even 7nm technology?
I came away from Semicon West with the impression that confidence in the demand for semiconductors over the next five years is strong, a reasonable view given consumer appetites for smartphones and tablets.
But the question over what node will be the entry point for 450mm became more urgent at the 2011 Semicon West. There may not be an answer to that question, as different vendors (Intel, for example) may be at the 9nm node when the major foundries are just starting 11nm risk production.
As the late songwriter Blaze Foley wrote in his classic song, Clay Pigeons: “I’m tired of runnin’ ’round lookin’ for answers to questions that I already know.”
That enigmatic thought may apply here.