
122 Cray T3E User’s Guide
interprocess communication, 62
interprocessor communication, 25
L
LAPACK, 9, 10, 33, 34, 114
latency, 25, 62, 114
level 1 cache, 23
level 2 cache, 23
library, 49
Libsci, 10, 33, 54, 106, 114
linear algebra, 34
Linear Algebra PACKage, 34
linear scalability, 108
linking, 32
local disk space, 20
local memory, 20–22, 24
logging in, 13
loop optimization, 43
loop splitting, 43
loop unrolling, 46
M
macros, 50
defining, 50
mailing list, 11
main memory, 24
make, 48, 88
Makefile, 88
makefile, 48, 88
malleable, 16, 33, 41, 53, 114
man, 18
matrix product, 54, 106
memory, 20–22, 24
hierarchy, 21, 24
references, 25
memory-bounded speedup, 109
message passing, 10, 62, 114
Message Passing Interface, 9, 63
metacomputer environment, 116
Metacomputer Guide, 11, 18
microkernel, 28, 114
microprocessor, 21, 22
MIMD, 21, 114
modules, 48
MPI, 9, 11, 33, 62, 63, 114
MPI_ALLREDUCE, 65
MPI_BCAST, 64, 65
MPI_Bcast, 66
MPI_COMM_RANK, 64, 65
MPI_Comm_rank, 66
MPI_COMM_SIZE, 64, 65
MPI_Comm_size, 66
MPI_COMM_WORLD, 64, 66
MPI_FINALIZE, 64, 65
MPI_Finalize, 66
MPI_INIT, 64, 65
MPI_Init, 66
MPI_INT, 66
MPI_INTEGER, 64
MPI_IRECV, 65, 67
MPI_ISEND, 65
MPI_RECV, 65
MPI_REDUCE, 64
MPI_Reduce, 66
MPI_SEND, 65, 67
MPI_SSEND, 65, 67
MPI_SUCCESS, 64, 66
MPI_SUM, 64, 66
MPI_WAIT, 65
MPN, 27
MPP Apprentice, 10, 95, 97
mpprun, 16, 18
Multi Purpose Node, 27
Multiple Instruction, Multiple Data,
21
N
NAG, 10, 37, 114
name, 45
Netlib, 114
Network Queuing Environment, 17,
81
network router, 22
nobl, 45
nobounds, 45, 47
node, 114
non-malleable, 16, 33, 38, 115
noreduction, 57
nosplit, 45
notation, 10
nounroll, 45
NQE, 17, 81, 115
further information, 87
number of processors, 8
numerical libraries, 10
numerical Internet address, 14
O
one-sided communication, 10, 70
operating system, 20, 21, 28
operating system nodes, 30
optimization, 42, 56
C compiler, 55
cache, 25, 44
Fortran 90 compiler, 42, 43
Kommentare zu diesen Handbüchern