Home
last modified time | relevance | path

Searched refs:cost (Results 1 – 11 of 11) sorted by relevance

/oneTBB/examples/graph/fgbzip2/
H A Dcompress.cpp291 UInt16 cost[BZ_N_GROUPS]; in sendMTFValues() local
402 cost[i] = 0; in sendMTFValues()
469 cost[0] = cost01 & 0xffff; in sendMTFValues()
470 cost[1] = cost01 >> 16; in sendMTFValues()
471 cost[2] = cost23 & 0xffff; in sendMTFValues()
472 cost[3] = cost23 >> 16; in sendMTFValues()
473 cost[4] = cost45 & 0xffff; in sendMTFValues()
474 cost[5] = cost45 >> 16; in sendMTFValues()
481 cost[j] += s->len[j][icv]; in sendMTFValues()
492 if (cost[i] < bc) { in sendMTFValues()
[all …]
/oneTBB/doc/main/tbb_userguide/
H A Dappendix_A.rst20 A more subtle cost is *cache cooling*. Processors keep recently accessed
35 evicted data, at the cost of hundreds of cycles for each cache miss. Or
40 Another cost is *lock preemption.* This happens if a thread acquires a
H A DAutomatic_Chunking.rst7 A parallel loop construct incurs overhead cost for every chunk of work
H A DContainers.rst35 Notice that highly-concurrent containers come at a cost. They typically
H A DMemory_Allocation.rst55 cost in space, because it must allocate at least one cache line’s
H A DControlling_Chunking_os.rst86 the cost of reducing potential parallelism. The overhead as a fraction
/oneTBB/doc/main/tbb_userguide/design_patterns/
H A DLazy_Initialization.rst23 The cost of operations that take place during the initialization
H A DDivide_and_Conquer.rst37 compared to cost of solving the subproblems.
H A DWavefront.rst124 consideration, because the atomic counting incurs a cost for each
/oneTBB/
H A Dthird-party-programs.txt221 …y used for software interchange, for a price no more than your reasonable cost of physically perfo…
/oneTBB/doc/
H A DDoxyfile.in1076 # cost of reduced performance. This can be particularly helpful with template