Index |
-, 1
Amdahl's Law, 1
block, 1
cyclic, 1
local, 1
array mapping, 1
backplane bandwidth, 1
bandwidth, 1
buffer congestion, 1
cache use efficiency, 1
Cluster Runtime Environment (CRE), 1
collocating processes, 1, 2, 3
-xalias, 1
-xarch, 1
concurrency, 1
cyclic message passing, 1
data movement, within a process, 1
deadlock, 1
dedicated access, 1
distributed-memory programming, 1
environment variables, summary of, 1
external cache, 1
full connections, 1
gprof, 1
interconnects, among nodes, 1
L2 cache, 1
latency, 1
lazy connections, 1
LD_LIBRARY_PATH, 1
libsunperf, 1
background, 1
defined, 1
Load-Sharing Facility suite, 1
mapping processes to nodes, 1, 2
memory bandwidth, 1
memory latency, 1
mpCC, 1
mpcc, 1
mpf90, 1
MPI profiling interface (PMPI), 1
MPI web page, 1
multithreaded MPI jobs, 1
nonuniform memory architecture, 1
pipelined messages, 1, 2, 3, 4, 5
problem size, reducing, 1
number of processes, 1
order, 1
rank, 1
rank sequence, 1
process grids, runtime mapping of, 1
MPI interface, 1
Solaris utilities, 1
profiling alternatives, compared, 1
programming practices, general, 1
remote shared memory (RSM), 1
S3L routines, summary of performance guidelines, 1
S3L Toolkit, 1
ScaLAPACK, 1
serializaiton, reduction of, 1
shared-memory array allocation, 1
shared-memory programming, 1
Sun Performance Library, 1
symmetric multiprocessor (SMP), 1
TCP connections, 1
TCP retransmission of data, 1
throttled communications, 1
timer calls, 1
VAMPIR, 1
Copyright © 2003, Sun Microsystems, Inc. All rights reserved.