A multi-chip module for physics experiments

MCMs are widely adopted as assembly solutions for multi-die based systems, where area, performance, and costs are critical constraints. This paper describes both the project strategies and production flow that are to be adopted to realize an MCM-D for data acquisition in high-energy physics experiments. The activity starts from the results of RD/16 CERN project, and is part of the LAP Esprit project. The paper details the most critical issues faced in the production phase, and analyzes how they influenced system partitioning and component design. Moreover, it presents the design-for-testability methodologies adopted at both chip and MCM levels to achieve low defect levels and high production yields, minimizing the overhead in terms of system performance and area occupation. This work should demonstrate the feasibility of the MCM technology in such high speed data processing systems, where both size and cost constraints are important. (10 refs).