and PV-DM on the CRAY C90 and with PV-DM on the Intel Delta and Paragon. CPU times per processor on the C90 are listed in Table 1. Times are given for both oneand four-processor runs with PVSOLVE. Linear speedup is obtained, since there was negligible load imbalance. The table also contains a single-processor time for PV-DM. This time is a factor of 2.7 slower than the single-processor time for PVSOLVE, which was optimized for the C90. Table 1 B- 99 Performance Results for PVSOLVE and PV-DM on CRAY C90: Mach 3.0 HSCT case with 88,404 equations Processors CPU time per Code processor (s) 1 140 PVSOLVE 4 35 PVSOLVE 1 376 PV-DM Table 2 Performance Results for PV-DM on Distributed-memory Computers: Mach 3.0 HSCT case with 88,404 equations Computer Processors Compute Send Receive time (s) time (s) time (s) Intel Paragon 128 24 12 82 (SUNMOS) 256 12 12 97 512 6 11 105 Intel Delta 128 36 119 261 256 17 101 282 512 9 100 287 Wall clock times for PV-DM on the Intel computer

Parallel computers offer the opportunity to significantly reduce the computation time necessary to analyze large-scale aerospace structures. This paper presents algorithms developed for and implemented on a massively-parallel computers hereafter referred to as Scalable High Performance Computers (SHPC) for the most computationally intensive tasks involved in structural analysis, namely, generation and assembly of system matrices, solution of systems of equations and calculation of the eigenvalues and eigenvectors. Results on SHPC are presented for large-scale structural problems (i.e. Models of high speed civil transport).