Electronic-based interconnections, whether chip-to-chip or between systems, are interently hierarchical. As these interconnects become large so does the total interconnect cross-sectional area. This leads to the formation of information bottlenecks. The future of computer systems will depend on the elimination of these bottlenecks and without question these systems will rely upon the use of photonics-based components. Fiber optic technology has the intrinsically high bandwidth needed to meet the future demands for high performance information processing. However, practical use of high-bandwidth photonics-based computers, and signal processors, will depend on the availability and acceptance of low-cost, high performance optoelectronics-based components. These components are necessary to fully develop the following key technology areas that are needed to make the transition from electronics to optoelectronics. The key technology areas are: hgih-speed optical memory to buffer data, optical header recognition and high-speed switching. The proposed research is to develop a novel, ultra-fast (Terabit/Photodetector), two dimensional photodetector matrtix array building block for "Photonics Information Processing." This matrix array is based on a novel Metal Semiconductor-Metal photodetector technology and a novel polymer waveguide-based photonic submount.
Keywords: Epitaxial Liftoff; Polymer Waveguide; Metal-Semiconductor-Metal Photodetector; Optoelectronic Matrix