Supercomputing provides large-scale computational resources for solving the largest and most challenging scientific problems, including issues like weather and protein folding simulations. Seymour Cray was a major influence on the architecture of silicon-based, electronic supercomputers. Modern supercomputers continue to be architected using silicon-based electronics and the designs suggested by Cray. However, several new architectures are emerging for computers that are not based on silicon electronics. These new architectures leverage biological, molecular, and quantum physics effects to perform computation in novel ways that promise to solve scientific problems that are intractable to silicon-based, electronic supercomputers. A review of the research literature on supercomputer architecture and emerging non-silicon computers reveals several potential research opportunities.
