The technological trends of micro- and nanoelectronics are mainly characterized by miniaturization, increasing levels of technology & function integration and eco-designing, while the business trends are mainly characterized by cost reduction, short-time-to-market. These trends together lead to increased chances and consequences of failures, increased design complexity, dramatically decreased design margins and increased difficulty to meet quality, robustness, reliability and shorter-time-to-market requirements.
Therefore one of the key aspects in electronic packaging is numerical prototyping including, which is capable of combining all those aspects into one design framework. However, within electronics industry, design and qualification are still largely depending on one's experience. Often, up to 10 cycles (material development/pre-selection, concept designing, building and testing multiple physical prototypes) are needed, with some qualitative support from numerical simulations. Quality, robustness, reliability and multi-scaling problems are usually dealt with physical prototyping, wherein reliability qualification testing with duration of 6 months is no exception. Clearly, this experience-based design and qualification method cannot lead to competitive design with short time-to-market, optimized performance, low costs, and guaranteed quality, robustness and reliability. This can be achieved by advanced and innovative numerical prototyping methods.
For numerical prototyping, "accurate and efficient prediction models" and "advanced simulation-based optimization methods" are the two core building blocks. By combining these two building blocks in a proper way, one can predict, qualify and optimize the physical behaviour and/or trends of novel products against the actual requirements prior to major physical prototyping, manufacturing investments and reliability qualification tests. A typical virtual prototyping procedure would consist of different numerical methods such as FEM (Finite Element Method), MM (Molecular Modelling), DOE (Design of Experiments), RSM (response Surface Analysis) and single or preferably multi-criteria Optimization.
The results of numerical prototyping can be used to predict, qualify and optimize the behavior and/or trends of micro- and nanoelectronics devices against the actual requirements prior to major physical prototyping, manufacturing investments and reliability qualification tests. It should be also noticed that traditional experiments and tests will continue to play an important role in the coming future: - firstly, they are needed in providing inputs for modeling, such as characterizing material and their interface behavior (material properties, damage initiation, evolution and failure criteria), - secondly, the accuracy of the developed numerical models and simulation results need to be verified experimentally in the whole range of the design spaces, and by covering all the critical processes parameters.