A study on the feasibility of using MCMs: a designer's perspective

Abstract

The purpose of the multichip module (MCM) project discussed was to develop and test a moderately complex, high-performance MCM, and in doing so, to determine the feasibility and constraints of using this approach to high-density hybrid integration. The end product was to be a working MCM prototype that could be characterized and tested at speed. The project was divided into four phases: (1) vendor evaluation; (2) electrical design; (3) mechanical design; and (4) reevaluation. Each phase is described. This work led to the determination that MCMs are a viable technology when speed and real estate are design drivers. It was found that maintaining good communications with the vendor throughout the design cycle is important. The limitations on the design tools may require the use of multiple tools to verify the functionality and performance of an MCM.<>