The Distributed Clearance Project: A Case Study in Distributed Computing

Distributed computing promises to break through computational performance barriers by cutting large problems into small pieces which can be solved by many machines acting in concert. However, not all problems are good candidates for this sort of "divide and conquer" strategy. Even large, compute-intensive problems that appear easily divisible on the surface may have dependencies on data and dependencies across partitioned units of work that may prove to negate the advantages associated with distributing the workload. The Distributed Clearance Project demonstrates both the potential and the pitfalls of distributed computing. The project seeks to take a very computationally intensive task (calculating part clearance tolerances in larg