Industry dynamics require a new look at risks and total cost
11/01/2004
Equipment Buying in Japan Part One: Outsourcing
As the semiconductor industry emerges from its most precipitous and prolonged slump in history, Japanese chip manufacturers are closely examining their business models to see what lessons can be learned that will permit them to survive, and even prosper, in the face of radical cyclical swings in the global market. A number of converging forces are making this examination all the more critical.
One is the move to 300mm wafers, and all the technical and economic consequences of this shift to a new mode of manufacturing. Another is the continuing excursion into the submicron realm of transistor size, from 130nm to 90nm feature sizes. Finally, there is the question of how to build a new manufacturing infrastructure to support the first two trends, but in a way that makes companies resilient in the face of cyclical demand swings.
Effects of wafer test and scaling
The move to 300mm wafer production obviously poses considerable risk in terms of capital exposure and is driving the use of joint ventures to build 300mm fab facilities. These new fabs require highly advanced wafer manufacturing and handling equipment, and demand large expenditures for equipment.
Consider wafer test. By more than doubling chip output, 300mm wafer manufacturing mandates a significant upgrade in test capacity, both at the wafer stage and the final test stage. With many devices/wafer, wafer test requires substantially more probe touchdowns/wafer, which significantly increases the number of test systems needed to ensure that manufacturing throughput is maintained. Over time, new probe technology that tests more devices/touchdown will become available, as well as techniques that make test vectors available to more devices in parallel. While technology and techniques like these upgrade capacity and reduce test resource requirements, they require considerable capital expenditures.
At the same time, the submicron migration to 90nm and beyond is substantially affecting capital equipment expenditures. An examination of the chip manufacturing process explains why. In microlithography and maskmaking, stepper wavelengths are descending from 248nm to 193nm, and will require advanced techniques such as phase-shifting and immersion lenses to reach the 45nm level. For CMP, the growing number of interconnects and the move to copper mandate new tools. Ultrashallow junctions in leading-edge devices require a new generation of ultralow-energy ion implanters. For deposition, the need to fill in very small features without leaving voids will require new equipment capable of atomic layer deposition (ALD). Finally, deep-submicron devices pose all sorts of challenges for assembly and packaging. This new generation of chips will often involve entire systems-on-a-chip (SoC), with very dense I/O arrays that routinely require flip-chip wafer-bumping processes.
Japan is one of the world leaders in capital expenditures with $9.7 billion planned for 2004, up from $7 billion the previous year. This represents 37.9% of the world total for this year. However, the business considerations involved in this current round of spending differ considerably from previous upswings in capital outlays. At the core of these considerations is a major shift not only in economics, but also in business philosophy.
Bucking tradition
Traditionally, Japanese companies have viewed direct ownership of production assets as a major source of corporate strength, and have maintained a highly proprietary view of equipment on the factory floor. Equipment was purchased, held until the end of its useful life, and then sold off almost as an afterthought. This view was reinforced by the industry's relative prosperity prior to the late 1990s, when large profits produced cash flows that in turn produced capital funds for aggressive expansion.
When IC sales plummeted into record negative growth in 2001, however, the amount of cash available for capital equipment expansion shrunk dramatically at the same time as idle production capacity put considerable downward pressure on the return on investment (ROI) for existing equipment. In this unprecedented environment, the traditional view of direct ownership of capital assets through outright purchase began to shift. It is now being replaced by a much more flexible view of capital equipment that is based on conserving cash traditionally spent on acquisitions to cover other areas of business operations, and thus preserve a company's strategic options throughout the inevitable fluctuations in the semiconductor marketplace. This new outlook sees each individual piece of equipment in terms of a continuum that covers its entire life cycle from the point of acquisition to eventual remarketing or reselling.
Armed with this more holistic approach to capital acquisitions and expenditures, management can consider a broad range of different strategies that meet production requirements while ensuring that cash flow is conserved. In addition to outright purchases, a mixture of loans, leases, and sale/leasebacks provide useful options. Loans provide the obvious benefit of conserving cash flow, and can be structured with payment terms that meet various business objectives. Leases reduce the amount of debt on the balance sheet and thus improve debt-to-equity ratios. They also provide a hedge against equipment that becomes prematurely obsolete. Sale/leasebacks are an arrangement whereby a company purchases a piece of equipment and immediately leases it back to the seller. This option keeps the equipment on the balance sheet, but provides steady cash infusion from the lease payments.
A strategically balanced acquisition plan takes the long view for any given piece of equipment and assesses its overall value by taking into account its eventual disposition in the global market for used equipment. This type of analysis requires extensive knowledge of the general flow of technological advance in semiconductor manufacturing systems, and how it will ultimately influence the disposition and replacement of each piece of equipment. It also requires a thorough understanding of current technology trends, trailing technology curves, and current market values.
At the outset of the equipment evaluation and selection process, it becomes critical to wait for the process to be efficiently and comprehensively completed, and not to move too rapidly for fear of falling behind competitors. Premature acquisition decisions may exclude valuable emerging technologies that offer significant competitive advantages. Each evaluation and selection process presents an optimum window of timing that balances thorough consideration with decisive action.
A balancing act
When equipment has reached the end of its useful productive life in a given fab environment, there must be an effective set of disposition policies that dictate how to remarket equipment in a manner that yields the best overall gain. Without these policies, a company will often spend excessive amounts of time, money, and energy to realize the true market value of a piece of equipment, which can considerably diminish the net return on the sale.
As an alternative, equipment can be marketed rapidly by pegging the sales price well below market value, which simplifies the sales effort but also reduces the net return on the sale. Resorting to equipment brokers, who profit by buying low and selling high, has the same effect.
A disposition strategy that produces maximum value from capital equipment requires extensive knowledge of both the local and global markets for the equipment in question, with potential sellers and buyers often on opposite sides of the world. If the seller confines itself to its immediate geographic region, it risks losing many high-value sales opportunities. However, to develop a global reach in remarketing requires extensive knowledge of international trade issues such as monetary exchanges, tariffs, local taxes, and safety regulations. So once again, the tradeoffs of gain vs. loss must be carefully assessed.
As business leaders within the Japanese semiconductor industry come to realize the need to explore alternative methods and attitudes regarding capital equipment, one major concern is the ultimate cost of adopting these new policies. A new focus on intensive management of the entire equipment life cycle diverts time and energy away from a chipmaker's core focus on making and selling semiconductors.
One reasonable and effective solution is outsourcing these functions to a third party that has already mastered the economic and technical intricacies of equipment lifecycle management. Such firms continuously monitor technology trends, fabrication, and testing and packaging processes. They also possess expert knowledge on all the details and relative advantages of leasing, purchasing, sale/leasebacks, equipment evaluation/locator services, and remarketing
eselling of equipment. Some also provide an array of financial packages designed to maximize the productive value of equipment while conserving cash expenditures on capital items.
For more information, contact Annika Hafner, VP of marketing, at GE Commercial Finance, Global Electronics Solutions; ph 858/320-7340, e-mail [email protected]; or visit GE Commercial Finance, Global Electronics Solutions on the web at www.GEelectronicsweb.com.