Abstract 7th MAPLD International Conference Washington, DC September 8-10, 2004 Title: Improved Long-term Reliability Evaluations for DoD Microelectronics Presenter: Vance Anderson Senior Engineer/Program Manager DoD - Defense Microelectronics Activity Microelectronics Systems Branch th 4234 54 Street McClellan, CA 95652 Phone: (916)231-1646 E-mail: firstname.lastname@example.org Introduction: The mission of the Defense Microelectronics Activity (DMEA) is to leverage the capabilities and payoffs of advanced technologies to solve operational problems in weapon systems, increase operational capabilities, reduce operation and support (O&S) costs, and reduce the effects of diminishing manufacturing sources (DMS). DMEA assists system managers by applying both available leading-edge technologies and innovative applied research and development (R&D) approaches to develop solutions to current problems. The organization translates microelectronics research technologies into solutions and evaluates their feasibility to address both current and impending microelectronics technology challenges for DoD. The DMEA is also the Executive Agent for DoD Integrated Circuit (IC) Microelectronics Diminishing Manufacturing Sources and Material Shortages (DMSMS). Microelectronics obsolescence is now the main factor driving DoD system degradation. The DoD increasingly relies on the use of "smart" weapons systems. The components that make these systems smart are the complex microelectronics devices that form the core of their functional capability. However, microelectronics technologies are extremely dynamic and generally become obsolete about every 18 months. Microelectronics obsolescence is a horizontal, technology-based issue rather than a vertical one, since systems throughout the entire DoD use the same or similar microelectronics devices. When a device becomes obsolete, every system using that device has a problem. As it is true of the problem, the solution must have the ability to cut across the many systems throughout the entire DoD As the Executive Agent, we evaluate the entire spectrum of microelectronics for current and future obsolescence issues, evaluate the feasibility of potential solutions, recommend solutions at all complexity levels, provide guidance, and recommend policy and procedure changes. Presentation Abstract: Defense Microelectronics Activity (DMEA), as the DoD Executive Agent for Microelectronics DMSMS, is concerned with the long-term reliability of modern microelectronics in DoD weapon systems. Of particular concern are programmable logic, microprocessor and memory structures due to their advanced designs and fine feature sizes. Many semiconductor manufacturers are designing products with reduced margins in order to gain higher performance in the consumer markets. This consumer market also has a different product lifecycle when compared to Department of Defense systems. The DoD demands high reliability over a much longer system lifetime when compared to consumer systems. DMEA has several initiatives to aid in the assessment of long term reliability of modern semiconductor structures. DMEA is working with industry and university researchers to develop a model-based algorithm to evaluate device and system reliability for a multitude of device types, foundry processes and feature sizes. The availability of a more accurate reliability model will enable the designer to make complete use of the devices performance without sacrificing system reliability. In addition, by introducing the operating parameters and environmental conditions into the reliability equation the system designer can tradeoff parameters such as performance and reliability. The models being developed are being validated by both actual field failure data and accelerated failure testing. The results, with a focus on programmable logic will be presented. DMEA is also exploring methods for evaluating semiconductor device reliability when models are not available. The DoD often needs to determine the reliability of a wide range of parts of which complex models simply do not exist. In some cases, device architecture and foundry process information is not even available. A novel method will be presented to determine device reliability for a variety of structures and architectures. Conclusion: DMEA is leading several initiatives to address long term reliability evaluations for FPGA, microprocessor and memory devices for the Department of Defense. These initiatives are leading to better tools and algorithms for both device and system designers. The efforts of these initiatives will be presented and their application discussed.