|
|
 |
|
|
|
|
occupancy, 367, 371 |
|
|
|
|
|
|
|
|
offered bandwidth, 361 |
|
|
|
|
|
|
|
|
offered request rate, 347, 360 |
|
|
|
|
|
|
|
|
offset, 41 |
|
|
|
|
|
|
|
|
on-chip caches, 299 |
|
|
|
|
|
|
|
|
one-one allocation, 129 |
|
|
|
|
|
|
|
|
open queues, 372 |
|
|
|
|
|
|
|
|
open-queue model, 365-376 |
|
|
|
|
|
|
|
|
operand value format, 33 |
|
|
|
|
|
|
|
|
operation set distribution, 144 |
|
|
|
|
|
|
|
|
optimum pipelining, 70 |
|
|
|
|
|
|
|
|
optimum static branch prediction, 493 |
|
|
|
|
|
|
|
|
ordering dependency, 460 |
|
|
|
|
|
|
|
|
orthogonal property, 21 |
|
|
|
|
|
|
|
|
out of order execution, 486 |
|
|
|
|
|
|
|
|
output dependency, 460 |
|
|
|
|
|
|
|
|
overflow, 129 |
|
|
|
|
|
|
|
|
overhead instructions, 112 |
|
|
|
|
|
|
|
|
overlapping the T cycle, 325 |
|
|
|
|
|
|
|
|
P-ratio, 112 |
|
|
|
|
|
|
|
|
PACK, 165 |
|
|
|
|
|
|
|
|
packed decimal format, 17 |
|
|
|
|
|
|
|
|
page mode, 353, 354 |
|
|
|
|
|
|
|
|
page size, 268 |
|
|
|
|
|
|
|
|
page table, 45 |
|
|
|
|
|
|
|
|
parallelism |
|
|
|
|
|
|
|
|
degree of, 526 |
|
|
|
|
|
|
|
|
finding in a program, 516 |
|
|
|
|
|
|
|
|
partitioning and scheduling |
|
|
|
|
|
|
|
|
effects of, 526 |
|
|
|
|
|
|
|
|
partitioning for multiprocessors, 512, 513 |
|
|
|
|
|
|
|
|
partitioning of address space, 358 |
|
|
|
|
|
|
|
|
peak rates, 376 |
|
|
|
|
|
|
|
|
perfectly colored pages, 325 |
|
|
|
|
|
|
|
|
performance, 83 |
|
|
|
|
|
|
|
|
linear model, 260 |
|
|
|
|
|
|
|
|
performance spike, 297 |
|
|
|
|
|
|
|
|
physical word sizes, 268 |
|
|
|
|
|
|
|
|
pipeline |
|
|
|
|
|
|
|
|
breaks, 158 |
|
|
|
|
|
|
|
|
dependency, 158 |
|
|
|
|
|
|
|
|
dependency effects, 196 |
|
|
|
|
|
|
|
|
interlock, 245 |
|
|
|
|
|
|
|
|
processor, 362 |
|
|
|
|
|
|
|
|
static and dynamic, 211 |
|
|
|
|
|
|
|
|
Poisson distribution, 88, 369 |
|
|
|
|
|
|
|
|
Pollaczek-Khinchine mean-value result, 373 |
|
|
|
|
|
|
|
|
precedence matrix, 462 |
|
|
|
|
|
|
|
|
prefetch, 269 |
|
|
|
|
|
|
|
|
prefetch register, 219 |
|
|
|
|
|
|
|
|
principle of inclusion, 304 |
|
|
|
|
|
|
|
|
procedure call, 150 |
|
|
|
|
|
|
|
|
process, 3 |
|
|
|
|
|
|
|
|
process addressing, 39 |
|
|
|
|
|
|
|
|
process management, 150 |
|
|
|
|
|
|
|
|
processor |
|
|
|
|
|
|
|
|
design, 3 |
|
|
|
|
|
|
|
|
early manufacturing, 106 |
|
|
|
|
|
|
|
|
framework, 186 |
|
|
|
|
|
|
|
|
performance constituents, 110 |
|
|
|
|
|
|
|
|
production phase, 107 |
|
|
|
|
|
|
|
|
ultimate cost, 107 |
|
|
|
|
|
|
|
|
program |
|
|
|
|
|
|
|
|
static size, 115 |
|
|
|
|
|
|
|
|
program counter, 6 |
|
|
|
|
|
|
|
|
program size, 37 |
|
|
|
|
|
|
|
|
program status word, 35 |
|
|
|
|
|
|
|
|
programming grains, 511 |
|
|
|
|
|
|
|
|
protocols |
|
|
|
|
|
|
|
|
broadcast, 538 |
|
|
|
|
|
|
|
|
directory-based, 538 |
|
|
|
|
|
|
|
|
snoopy, 538 |
|
|
|
|
|
|
|
|
putaway, 66 |
|
|
|
|
|
|
|
|
Q factor, 276 |
|
|
|
|
|
|
|
|
queue properties, 372 |
|
|
|
|
|
|
|
|
queueing models |
|
|
|
|
|
|
|
|
selection of, 394 |
|
|
|
|
|
|
|
|
queueing theory, 365 |
|
|
|
|
|
|
|
|
R+M instructions, 10 |
|
|
|
|
|
|
|
|
R/M instructions, 10 |
|
|
|
|
|
|
|
|
RAID, 655 |
|
|
|
|
|
|
|
|
random replacement (RAND), 286 |
|
|
|
|
|
|
|
|
RAS, 352 |
|
|
|
|
|
|
|
|
Rau's model, 365 |
|
|
|
|
|
|
|
|
rbe, 91 |
|
|
|
|
|
|
|
|
real address, 45 |
|
|
|
|
|
|
|
|
real data type, 15 |
|
|
|
|
|
|
|
|
referencing behavior, 361 |
|
|
|
|
|
|
|
|
register |
|
|
|
|
|
|
|
|
bypassing, 241 |
|
|
|
|
|
|
|
|
register allocation, 129 |
|
|
|
|
|
|
|
|
register bit equivalent, 91, 132 |
|
|
|
|
|
|
|
|
register fetch, 65 |
|
|
|
|
|
|
|
|
register output gating, 7 |
|
|
|
|
|
|
|
|
Register plus Memory architecture, 11 |
|
|
|
|
|
|
|
|
register windows, 129 |
|
|
|
|
|
|
|
|
Register/Memory architecture, 11 |
|
|
|
|
|
|
|
|
relative call depth, 128 |
|
|
|
|
|