In 2002, Dr. Prabasaj Paul published an iterative algorithm for bandgap maximization in photonic crystals (P. Paul and F. Ndi, e-print physics/0205091). The computational requirements of this algorithm are such that typical personal computers and high-level languages such as Mathematica result in performance characteristics less than those needed to reach the numerical accuracy requisite for fabrication.

During summer 2003, the algorithm was reimplemented for parallel processing in C++, thus making higher resolutions tractable. Data rendered at these resolutions ended up providing support for two hypotheses presented in the aforementioned paper.

From that time, further optimization and parallelization have ensued. Tractable resolutions are now sufficient to further explore other aspects of the problem, such as iterative convergence criteria. Work is presently underway to modify and deploy the parallel solver to utilize otherwise inactive computational resources—campus computer labs, for example.