Vector Processing on CRAY-1 and CRAY X-MP
Vector processors provide great performance improvements in scientific and engineering computations. This is achieved by a specialized architecture designed for the fast processing of certain types of operations. On the other hand this specialization in hardware and architecture leads to a corresponding specialization in programming and program optimization for such machines.
KeywordsFunctional Unit Vector Length Execution Sequence Vector Element Clock Period
Unable to display preview. Download preview PDF.
- S.S. Chen, Large-scale and high-speed multiprocessor system for scientific applications: CRAY X-MP series, in: “High-Speed Computation,” Vol. F7, NATO ASI series, Springer-Verlag Berlin, Heidelberg (1984).Google Scholar
- CRAY X-MP Computer Systems, CRAY X-MP Series Mainframe Reference Manual, HR-0032, CRAY Research Inc.Google Scholar
- P.W. Schwarztrauber, A parallel algorithm for solving general tridiagonal equations, Mathematics of Computation, Vol. 33, No. 145 (1979).Google Scholar
- F. Hossfeld, Parallele Algorithmen, in: Informatik Fachberichte, Band 64, Springer Verlag (1983).Google Scholar
- W. Oed, and O. Lange, Transforming linear recurrence relations for vector processors, in: Parallel Computing 83, Elsevier Science Publishers-North Holland (1984).Google Scholar
- V. Aho and J.D. Ullman,”Principles of Compiler Design”, Addison-Wesley Publishing Co. (1978).Google Scholar
- U. Detert, and H.W. Homrighausen, Analysis of the dynamic behaviour of FORTRAN programs, in: Proceedings SEAS Spring Meeting 1983, SHARE European Association, Nijmegen (1983).Google Scholar
- U. Detert, Performance comparison for CRAY-1/S and CRAY X-MP by means of FORTRAN kernels and user programs, in: Proceedings CRAY User Group Meeting, Paris (1984).Google Scholar