Index
367
Single-chip cloud computer (SCC), 9

Single error correction (SEC), 220

Single-event transient (SET) pulse, 201

Single-event upset (SEU), 200

Soft error rate (SER), 201

Soft errors, 199–203

in back-to-back inverter, 200, 200f

classification of effect, 201

correction, 206–210

combinational logic, 207–210,
208f, 209f
in latches, 207f
using time-shifted output, 209f
in D-type latch, 200, 201f

Sort-based mapping (SBMAP), 139

STALL/GO flow control protocol, 44, 44f

Static voltage scaling (SVS), 160

Store-and-forward (SAF) packet
switching techniques, 29

Subthreshold leakage current (I sub ), 156

Swap operation, 259–260

Switch allocation, 64

Switch arbiter (SA), 58, 205

Switching power consumption, 155–156

Switching techniques, 29–30

and packet format, 53–54, 54f

Switch-to-switch flow control
schemes, 218

System-level floorplanning, 268–271
constraints, 269–270
mesh topology constraints,
270–271, 270f

objective function, 269

variables

dependent, 268–269

independent, 268

System-level power reduction, 172–188

DFS, 179–185

characteristics, 184t

DFS algorithm, 183

history-based DFS, 181–182, 182f

link controller, 183–184

DVS, 172–179

characteristics, 173

components of links, 173f

hardware implementation, 178, 178f

history-based DVS, 174–178

runtime power gating, 186–188, 188f
VFI partitioning, 185–186
System-on-chip (SoC), 1, 75, 129, 191, 235,
319, 321

categories, 1

hand mapping of cores in four-

layered 3D, 349, 349t
integration and its challenges, 1–2
to NoC, 3–5
T
TD factor, 202

Template-based efficient mapping
(TEM) algorithm, 139

T-Error flow control protocol, 44, 45f

Test access mechanism (TAM), 235

Three-dimensional (3D) NoC, 317

Three-wire model

energy obtained from HSPICE, 82t

parasitic capacitance, inductance,
and resistance of, 81t

Throughput, 78

vs. locality factor, 85–86, 87f, 98–99,
98f, 109–110, 110f, 341–342, 341t

vs. offered load, 104, 105f

Through-silicon vias (TSVs), 8, 317, 319

misalignments, 322, 322f

Time division multiplexed access
(TDMA), 138

Time division multiplexing (TDW), 354

Topology graph, 120, 122

Topology reconfiguration, 304–311,
308f, 309f
architecture, 306–311
application mapping, 307–308
core-to-network mapping,
309–310
and route generation, 310–311
routers wrapped by
switches, 305f

logical, 306f

modification around routers, 305

multiplexer-based implementation,
307f
Traffic modeling, 81–84
Transient fault controlling techniques
inter-router link error control,
210–221
capacitive crosstalk avoidance
techniques, 210–216
Précédent

- 386/388

Suivant