Stategic Alliances for highly efficient 300 mm Waferfabs

State-of-the-Art waferfabs require a growing demand for integrated and highly efficient system solutions in terms of investment costs, running costs and yield improvement. It will be explained which factors are determinant to the profitability of a waferfab. Making the change from 200 mm to 300 mm wafers together with the jump to 0.25 μm technology will be the challenge the entire semiconductor industry within the coming years. The author explains why the jump to 300 mm is so critical in terms of technology and efficiency. The investment costs which are necessary to cope with the 300 mm challenge are more than any single company can handle. This is why strategic alliances have to be founded. The impacts are obvious: Systems or turnkey supplier to the semiconductor industry have to contribute as well as the chipmanufacturers to ease the burden of huge investment costs. 1. Growing Demands of the Industry in Terms of Investment Costs, Running Costs and Yield Improvement Investment costs as well as running costs of a fab increase tremendously. See fig. 1. Fig. 1 IC Facility Costs Page 1/7 2. The 300 mm Transition At Semicon/Japan 1995 the world s IC-producing nations agreed that 300 mm should be the next generation of silicon wafer size. At that time the production of these larger diameter wafers was expected to start in 1998. The conversion to 300 mm wafers is expected to lead to a significant reduction in manufacturing costs, improve yields and enable other productivity improvements. Regarding to six major factors the start up for production in 1998 seems to be too optimistic: Larger development times, extended wafer life cycles, market situation, materials shortages, technical obstacles and funding issues. The transitions to 150 mm and 200 mm respectively were funded in both cases by one single company. Regarding to the increased huge investment and developing costs no single company is willing and able to pay for the transition. The development time for a new wafer generation has increased significantly (see fig. 2). Fig.2. Wafer Development Time It has been estimated that it will cost the global semiconductor industry and its suppliers as much as US$ 14 billion to develop a 300 mm capability by the end of the decade. So, why is the industry making such an enormous effort ? The transition into 300 mm technology is merely driven by economics. The number of dies per wafer will increase by the factor 2,4 to 2,7. Big logic devices require bigger wafer diameter and thus increase die per wafer efficiency. See fig.3. For large CPU chips with 5 to 10 M transistors the efficiency on 300 mm wafers is higher. See fig.3. The cost per die is expected to be 25 40 % lower. Finally there are fewer fabs necessary to be built to meet the demand for chips. The 300 mm transition is going to increase investment costs and running costs for waferfabs. On the other hand, chip productivity increases with respect to the wafersizes.