Professor Mou is currently developing an optimizing compiler for his parallel language Divacon. The optimizer is directed by a built-in micro-analyzer and can perform semantics-preserving transformations. One of his ongoing research project is the study of the communication patterns in recursive parallel algorithms and their optimal mappings to mesh and hypercube architectures. He is also working on the design, implementation, and analysis of parallel algorithms for a number of problems in scientific computing.
In addition to the above, Professor Mou is interested in functional parallel programming languages and related issues. These include abstract interpretation, partial evaluation, and transformations of functional parallel programs. His recent research interests also include distributed data structures and parallel evaluation of recurrences.
Divacon: A Parallel Language for Scientific Computing Based on Divide-and-Ccnquer. 1990 IEEE Symposium on the Frontiers of Massively Parallel Computation, pp. 451-461.
Optimal Mappings of m Dimensional FFT Communication to k Dimensional Mesh for Arbitrary m and k (with X. Wang). 1993 Conference on Parallel Architectures and Languages Europe (PARLE93), Lecture Notes in Computer Science 694, pp. 104-119.
Spectrum Analysis of Communication Networks on Parallel Computers. Practical Aspects of Parallel Computing (ed. L. Tao), pp. 123-140. International Academic Publishers, 1994.