The research objective of this award is to develop a design framework that enables the modeling of multi-physics variability in 2.5D and 3D integrated micro-systems with a focus on silicon vias (TSVs), by accounting for both local and global variations across the system. Using this framework, methods will be developed for localized parametric optimization across the thermal-electrical-mechanical multi-physics domains for minimizing variability and maximizing performance. The difficulty in modeling these multi-physics interactions between the three domains is complicated due to the multi-scale dimensions of the structures involved and manufacturing variations. In addition, use of technologies such as micro-fluidics for cooling such systems further complicates the domains that need to be modeled. The deliverables include the development of a modeling framework for simultaneous analysis of the thermal, electrical and mechanical equations across a 2.5D/3D sub-system or system containing TSVs, and a means to enable the capture of the interaction between the domains and parameterization in critical areas for the cross domain optimization of the thermal, electrical and mechanical behavior.
If successful, this research will lead to a design and modeling framework that could address the problems being faced by the semiconductor industry for the design of 3D integrated systems. The results of this research will be disseminated to the design community and also used to develop curriculum material for 3D system integration to benefit both undergraduate and graduate students. High school students and teachers will also be educated through this award by creating awareness amongst them on the issues and challenges being faced by the semiconductor and electronics industry for miniaturizing electronic systems.