The semiconductor capital equipment industry has still not recouped $14 billion in investment in 300-mm wafer processing technology and is not prepared to develop 450-mm tools.
Editor's note: This article is part two of a two-part analysis on the state of the front-end capital equipment industry and its future prospects. Part one can be found here.
The implications of several semiconductor industry factors that I detailed in part one of this series of articles are far-reaching. With the industry in a difficult situation, with little product innovation from the rest of the industry and relatively poor financial performance given 300-mm product pricing, there is insufficient discretionary funding to develop a 450-mm generation of products to further the cause of Moore's Law.
The industry is still reeling from its $14 billion investment in 300-mm, for which returns are still not achieved. The performance of these stocks has underperformed the market and technology companies since 2006. Therefore, financial markets are not going to be ready to provide the financing and capital required to develop the 450 mm generation. The companies do not have the financial capability to develop the next wafer size generation.
Developing a new wafer size generation is an expensive process requiring considerable expertise. The core competencies of the front-end equipment company are the ability to develop integrated tools combined with process development capability, materials qualification, software and factory automation. These competencies are a significant barrier to entry particularly in a wafer size change since the entire process must be repeated as materials scaling from prior generations (in this case 300-mm) is not a simple task.
The current front-end companies are unique in this capability and it's not likely that the semiconductor manufacturer—irrespective of size—is going to be able to develop these skills sets any time soon. There are few front-end equipment suppliers that are multi-product such as Applied Materials Inc. and Tokyo Electron Ltd., and it is unlikely that even these two giants of the industry can spend the money and resources required to develop the next wafer size generation.
The behavior of the largest chip makers—Intel Corp., Samsung Electronics Co. Ltd., Taiwan Semiconductor Manufacturing Co. (TSMC) and I would count Toshiba Corp.—has had a lot to do with this problem in addition to the behavior of the equipment companies themselves. It seems to me that if the equipment companies cannot earn adequate returns, the biggest chip makers may have to go back to building and developing the equipment themselves—something that hasn't been done in the industry since the late 60s and 70s, when it was done by Texas Instruments Inc.
Samsung and TSMC have only been factors in the market since 1994. These companies do not have sufficient skills within their companies to develop equipment. Much of what they know about process development capability has resulted from their relationships with the equipment companies that have served them since their founding.
Samsung continues to sponsor efforts by insisting on Korean supply chain control and starting companies in Korea to build equipment. None of these endeavors have yielded a competitive product for consumption on the world stage.
The game is going to have to change and unfortunately it might change with the development of the Chinese equipment industry which today is in its infancy. I believe that the Chinese will support a homegrown equipment industry to support their internal semiconductor industry. One equipment startup, AMEC, has developed a state-of-the-art oxide etcher that has penetrated a number of companies in the global semiconductor customer base. This product development has taken place over a period of years and reflects significant investment and patience by China after initial funding by U.S.-based venture capitalists. The type of patience and persistence demonstrated by the Chinese in this startup is significant as most equipment startups in the last five years outside of China have resulted in failure.
The threat of a double-dip recession would potentially put the semiconductor industry in another situation of over capacity. Customers have learned a lot over the past few years about how not to have to buy new tools unless they are absolutely necessary. The 300-mm generation of tools has reduced the unit demand placed on the market.
Is 450 mm a business opportunity for US companies? Who will reap the benefits of 450 mm? Who will pay for the development? It is clear that we need fresh leadership from the equipment companies and the IC companies to properly fund and execute this project. Who is going to take the leadership?
Do IDMs and foundries want to bet their future on equipment made in the PRC? We have seen what the Chinese government is willing to do with supply chains (rare earths) if they perceive a political or economic advantage. My guess is that a Chinese equipment base may serve a domestic niche but any large scale market share gains at the expense of the major OEMs is unlikely.