High-performance spectral simulation of turbulent flows in massively parallel machines with distributed memory

Here we have demonstrated the possibility of very high performance in the implementation of a global spectral methodology on a massively parallel architecture with distributed memory. Spectral simulations of channel flow and thermal convection in a three-dimensional Cartesian geometry have yielded a very high performance - up to 26 Gigaflops per second on a 512 node CM5. In general, implementation of spectral methodology in parallel processors with distributed memory requires non-local inter processor data transfer, which are not restricted to be between nearest neighbors. In spite of their increased communication overhead, better performance is possible in global methodologies owing to their dense matrix operations and organized data communication. In this paper we outline a general methodology for the data parallel implementation of spectral methods on massively parallel machines with distributed memory and following the steps presented here very high performance can be obtained on a wide variety of massively parallel architectures.National Science Foundation 94/0