4 pages matching pipelined arithmetic units in this book
Results 1-3 of 4
What people are saying - Write a review
We haven't found any reviews in the usual places.
2nd Edition SUPERCOMPUTEVG achieve algorithm Amdahl's Law array of numbers bank conflict Cache memories cache or register clustered system compress/index Computational science contiguous cooling technology cycle Dhrystone ELECTRICAL AND ELECTRONICS ELECTRONICS ENGINEERS execution extended random access fetch and store Fortran full vector global memory global network granularity high perform hypercube index vector Informal Glossary INSTITUTE OF ELECTRICAL instruction set iterations least one CPU loop unrolling MFLOPS rate microcode MIMD Multiple CPU architecture multiprocessing multiprocessor multiprogramming multitasking number of non-zeros number of processors operations per second overhead overlapped functional units Parallel Computers parallel processing percentage vectorization performance improve pipelined arithmetic units point operations primary memory proc process multiple vectors puter memories random access memory Scientific Supercomputer Subcommittee SIMD single-instruction SISD speedup stream—multiple-data stream stride sub a half supercom SUPERCOMPUTEVG An Informal term throughput tion true ratio usually vector computers vector instructions vector length vector operation vector with relatively Whetstone