Executive Summit: Industry Leaders See a Promising but Challenging Future
Megatrends Pointing Upward—but Scaling, EUV, 450 Remain Major Hurdles
“The challenge with all these technologies is that we cannot decouple them from the economics,” said Rick Wallace, president and CEO of KLA-Tencor during the annual Executive Summit at SEMICON West. While aligning the economics of new production technologies such as EUV, III-V materials, FinFETS, and 450 mm wafers have yet to be fully worked out, the insatiable appetite for semiconductors by a rapidly-growing world middle-class promises a bright future for the semiconductor supply chain, according to a panel of equipment and materials industry executives.
Moderated by Jonathan Davis, president of the semiconductor business unit at SEMI, the annual Executive Summit brings industry leaders together to address the immediate and long-term outlook for semiconductor manufacturing, while directly confronting the most controversial issues in the chip ecosystem. Joining Wallace on stage were Terry Brewer, founder and president, Brewer Science; Doug Neugold, Chairman, CEO and president of ATMI; and Steve Newberry, CEO and vice chairman of Lam Research.
All the panelists were enthusiastic about the long-term growth prospects for semiconductors and the supply chain. “Electronics products have become not just a luxury, but a necessity,” said Newberry. “I don’t think we are anywhere near the top of the power curve. I think we are just approaching it.”
“The trend that I focus on is the ubiquitous role of semiconductors in our lives today, the two billion consumers who are entering the middle class in developing regions, and their insatiable demand for information and mobility,” said Wallace.
“You look at the mega-trends and they do nothing but reinforce my belief that computing power and memory capability are going to enable things that are even more amazing,” echoed Neugold. “They are going to address the major concerns that people have in the world: energy, health care, and more.”
Complexity and Cyclicality
While long-term growth seems assured, managing complexity and cyclicality remain the top challenges for the industry. “Complexity is going to keep on increasing,” said Brewer, “and the ability and manage complexity is going to be a part of how to be successful.”
“The same products, the same companies, the same approaches that drove the success of the electronics in the past will not drive it in the future,” said Newberry. “It will take a different tack. The drivers will be different.”
With long-term demographic trends appear favorable to chip demand—and external macro economic forces beyond their control—the panelists were surprisingly unconcerned about immediate sales trends and cycles. Complexity and long-term product development require a longer term outlook.
“You can’t manage the business short-term because the development cycles are too long,” said Wallace. “In our company, we are making investments in new products that won’t make an impact for five years. You can’t really modulate your investment over the swings that we see.”
“We are going to have rolling 12-month cycles and you have to invest through that,” said Newberry. “To me, it’s not about , it’s about 2013, 2014, 2015. It’s about secular demand trends for electronic products… and most of us don’t worry about the next quarter or two.”
One of the emerging ways that cyclicality and complexity is changing the business is through how companies collaborate with their customers. “Complexity forces you to think through structural issues,” said Neugold. “Very different discussions have occurred over the last three years to try to figure out a new collaborative model.”
EUV, 450 Timetable, Process Uncertain
In moderating the panel, Jonathan Davis pointed out there remains a high degree of technological uncertainty in the industry. Despite highly-evolved industry collaboration processes such as the ITRS, SEMI Standards, and the growth of international research consortiums, major process milestones in lithography, wafer sizes, and new transistor architectures remain unclear and have made coordinated industry planning more challenging and riskier than ever.
Extreme Ultraviolet (EUV) lithography rollout is becoming increasingly complex.
Throughput limitations on current tools have the industry rethinking, or reengineering, EUV implementation. According to Newberry, “I think EUV is ultimately going to be necessary. Today, you have a lot of activity really trying to solve some of the alignment and registration issues around quadruple patterning… semiconductor manufacturers and designers are going to limit EUV’s required presence to the fewest numbers of steps… Don’t be surprised if EUV gets pushed out.”
The timing, process and funding of an industry move to 450 mm wafers are still uncertain, even according to these insiders.
“You are not going to 450 for performance reasons; you are not going to 450 for perceived economic reasons,” said Newberry. “That’s what we are going to determine in the next three years. What are the real economics around what it takes to develop a suite of equipment and materials that actually can process at 450? That exploration is going to determine whether it occurs and at what speed it occurs.”
Rick Wallace agrees on the challenge of fully transitioning to 450, but indicates the timetable may occur faster than anyone realizes. “The infrastructure associated to be ready for [450 mm wafers] is going to take an entire supply chain and it is not nearly ready.” But he added, “We’ve seen initial orders for 450 systems at our company that we will deliver in the next 12-months…I think it  will happen.”
Newberry isn’t so sure. “If you think three to five years out, we’re going to be right in the middle of trying to go to volume production in 3D NAND, we will be dealing with vertical DRAM, we may even be dealing with MRAMs, we are certainly going to be right in the middle of FinFETS at 14 nm. We may even be dealing with III-V type structures. And if you think about all that complexity, of those new designs, new architectures and the scaling required at 14, 10 nm, and you want to put 450 on top of that?”
Wallace remains confident, however, that the industry will find a way to achieve the productivity gains that have consistently marked the history of semiconductors. “I think the way the industry is trying to go about 450 is significantly different and better than 300… I do think it’s a better process… These are initiatives that are far more thoughtful because the stakes are higher.”
“The issue is how to collectively move — the semiconductor manufacturers and the equipment people — and how do we do a lot of early work together and not have five different pilot lines…” Wallace continued, “I think somebody funding it beyond the equipment companies on their own will be necessary and the economics have got be shared.”
August 2, 2011