Implementations of Parallel Algorithms

Here I'm interested in getting demonstrably good performance from highly parallel computers on key problems. Most of this work was carried out on MasPar MP-1 and MP-2 computers, but the techniques apply to other architectures as well. Here are some references, and, in some cases, access to publications in postscript form and program sources.

Sorting

Connected Components

Rendering

Interactive Simulations

Load Balancing

Lattice Gas and Lattice Boltzmann Automata

Fast Multipole N-body simulations

prins@cs.unc.edu