The inviscid gas dynamics equations have been solved on the nCUBE 2s computer to better understand the challenges and opportunities that parallel processing presents. Two basic computational fluid dynamics (CFD) application problems representing different flow physics are used as test cases: supersonic internal flow and subsonic flow over an airfoil. Analysis of the parallel performance for these applications is presented following a fixed time measurement approach. Results from this study demonstrate that even though these flow simulation problems are communication-intensive, they are scalable. Parallel computing is shown to be an effective tool for future CFD challenges.>