ГОСТ Р 57700.27-2020. Национальный стандарт Российской Федерации. Высокопроизводительные вычислительные системы. Термины и определения

Алфавитный указатель эквивалентов терминов

на английском языке

accelerator	36
anti dependence	78
archive	19
arithmetic accelerator	37
average distance	50
bandwidth	91
barrier synchronization	81
batch job	63
benchmark	86
bisection	47
bisection bandwidth	98
bisection width	99
communication channel	43
computing node	29
computing subsystem	12
computing system	2
connectivity	96
cooling system	27
core	32
critical section	82
data dependence	76
data parallelism	74
decomposition	70
diameter	100
disk array	53
distance	49
efficiency	104
engineering systems	25
file subsystem	18
file system	51
Flynn's taxonomy	20
full bisection	97
functional decomposition	71
functional node	10
functional parallelism	75
general-purpose high-performance computing system	8
general-purpose processor	31
geometric decomposition	72
granularity	106
graphics processing unit (GPU)	38
hardware	28
heterogeneous high-performance computing system	7
high-performance computing (HPC)	1
high-performance computing system (HPCS)	3
high-performance computing system architecture	5
homogeneous high-performance computing system	6
hop	45
hot replacement	55
hot reserve	56
hybrid computing node	33
hybrid parallelization model	67
interconnect	39
interconnections subsystem	16
interconnections subsystem topology	44
just a bunch of disks (JBOD)	54
latency	90
link	42
load balance	107
load balancing	84
load disbalance	108
management and administration subsystem	13
mapping	85
message passing interface (MPI)	68
message passing parallelization model	65
MIMD (Multiple Instruction stream, Multiple Data stream) architecture	24
MISD (Multiple Instruction stream, Single Data stream) architecture	22
multiprocessor computer	4
node	46
NUMA architecture	35
occupied space	95
OpenMP	69
output dependence	79
parallel computing	61
parallel file system	52
performance	87
port	41
power consumption	94
power efficiency	93
power supply system	26
power usage effectiveness (PUE)	92
process	58
process's context	59
processor (CPU)	30
race condition	73
real performance	89
resource manager	64
route	48
scalability	105
serial computing	62
service subsystem	15
shared memory parallelization model	66
SIMD (Single Instruction stream, Multiple Data stream) architecture	23
SISD (Single Instruction stream, Single Data stream) architecture	21
SMP architecture	34
special-purpose high-performance computing system	9
speedup	101
storage subsystem	17
strong scaling method	102
subsystem	11
switch	40
synchronization	80
system software	57
theoretical peak performance	88
thread (Light Weight Process)	60
true dependence	77
user access subsystem	14
vectorizing	83
weak scaling method	103

УДК 001.4:004:006.354

ОКС 35.240.50

Ключевые слова: высокопроизводительная вычислительная система, коммуникационная подсистема, архитектура вычислительной системы, вычислительный узел, процессор, модель распараллеливания, процесс, поток, коэффициент ускорения, эффективность распараллеливания