As computing environment has been developed during last half a century, many difficult mechanical problems has been solved. But when they have more complex geometries or we want to search more profound meanings of the physics, then faster and larger memory computers are needed. While the main architectures of computers are designed by using a single CPU in a computer until now, but we can't solve the problems as mentioned above. To overcome these limits parallel computers have been developed.
Two different architectures parallel computer, the SIMD type MasPar and the MIMD type Cray, are used to parallelize the Euler Code. In the Euler Code the finite volume nodal scheme for unstructured grid is adopted due to its fast and stable behavior. The directive-like method and ForTran 90 are used to parallelize in the Cray and the MasPar machine respectively. For the transonic calculations over NACA 0012 airfoil, algorithmic speed-up ratio, total CPU time, and run time are computed. As the number of CPUs increases, larger speed-up ratio can be achieved.
