Collaboration, Coordination And Consolidation
By Ed Sperling
By anyone’s standards, the ability of the semiconductor industry to reduce costs is the pinnacle of industrial automation. The dollar that bought a single transistor 30 years ago will buy 1 billion transistors today, an efficiency gain that has put powerful computers into the pockets of billions of people and transformed how we interact with the rest of the world.
But a combination of increasing complexity, shrinking margins and new ways of making chips also are beginning to change the foundation of the semiconductor equipment business, Mike Splinter, chairman and CEO of Applied Materials, told a packed ballroom of semiconductor executives today at the SEMI Industry Strategy Symposium. He said that advanced transistors, interconnects and patterning, as well as the move to wafer-level packaging, will have a fundamental impact on every aspect of the IC business.
“Change is accelerating,” said Splinter. “Compared with the last 15 years, the next five years will have more changes and more inflection points. And it’s not just about complexity. It’s happening at the foundational level of how an IC is made.”
Even the transistor itself is changing from 2D to 3D structures, where current leakage can be controlled more effectively. Coupled with ultra low-k films and improvements in reliability and resistivity in the interconnects, an increase in the number of steps needed for patterning, the move to wafer-level packaging instead of die-level packaging, and the move to 450mm wafers and the industry will have to shift even further.
Splinter said collaboration across the ecosystem will need to be done earlier and at a deeper level. It also will have to focus heavily on high-value problems, where return on investment in R&D is highest. And it will require open innovation practices so that some of the IP can be opened up to share more useful information and to reduce waste.
“There are many new materials that don’t fall into our core expertise,” he said. “We also need to figure out how we are going to do this change to 450mm better than the 300mm transition.”
He estimated the cost of shifting to $450mm will be between $15 billion and $20 billion, compared with about $12 billion for 300mm. But the 300mm transition could have been done for less, too. “We had a number of false starts. We can avoid that and make it more efficient, but to do that we will need a forum for equipment suppliers to talk to semiconductor suppliers so they can have an open dialog. And we need to work together on the availability of government funding.”
Consolidation has been a recurring theme at this year’s ISS. The acquisition of Novellus by Lam Research and the acquisition of Varian by Applied Materials are enormous purchases. Applied paid $4.9 billion for Varian last year, while Lam will pay $3.3 billion for Novellus. That raises the stakes not only for jumping into this market, but it also is an indication that even established large companies are ceding ground to their bigger rivals.
The numbers seem to support this. Applied Materials numbers show that in 2000 54% of the equipment market was held by the top five companies. In 2012, the number has risen to 67% of the market. Splinter said the consolidation is just another example of the need to drive waste out of the industry. But even that trend is accelerating. In 2010 only 61% of the market was held by the top five equipment suppliers.
Tags: Applied Materials, SEMI