Visualizing staggered fields and analyzing electromagnetic data with PerceptEM

In order for Moores law (exponential growth of compute capability) to continue to hold, the bandwidths of interconnects will need to be increased by 10x or moreto well beyond 100Gbs. This represents a formidable challenge to Integrated Circuit (IC) manufacturers because higher frequency will exacerbate already strained power dissipation of traditional copper interconnects. As a result, server and networking equipment companies (IBM, CISCO, Lucent/Alcatel) are pushing the semiconductor industry toward investing in using light (photonics) to propagate information. We will improve the design productivity of silicon photonic IC components, which target Silicon on Insulator (SOI) manufacturing processes through the development of automation and modeling tools packaged with a set of pre-designed parametric silicon photonic IC passive building blocks (waveguides, couplers, resonators, modulators, and Y-junctions). The geometric parameters of these blocks can be modified and simulated to predict performance for a specific target SOI manufacturers process. The ability to accurately model and simulate specific instances is enabled by our unique three-dimensional physical modeling, multi-physics simulation and cluster computing capabilities. In Phase I, we laid down the groundwork for visualizing and analyzing staggered fields, which are common in electromagnetic simulations. In Phase II, we implemented conservative algorithms to compute line and area integrals of staggered fields; the same algorithms will be applied in Phase IIB to characterize silicon photonics components. The outcome of the work performed in Phase I and II will be applied to characterize optical components with unprecedented accuracy using an S-matrix formalism. The result will be that each optical component can then be integrated into a larger circuit using tools and processes similar to those currently employed in electronic integrated circuit design. Most silicon photonics design tool and manufacturing sourcing (e.g. Luceda and Phoenix Design) is done outside of the U.S., while the end use for this technology within big data companies (Apple, Amazon, Google, Facebook, Microsoft) and their equipment suppliers (CISCO, IBM, Dell, HP, Oracle) are U.S. companies. We believe the automation of the key building blocks used to design and specify the manufacture of silicon photonic ICs will enable these US companies and their IC partners to accelerate their ability to design silicon photonic ICs.