Home
last modified time | relevance | path

Searched refs:parallelism (Results 1 – 25 of 29) sorted by relevance

12

/oneTBB/doc/main/tbb_userguide/
H A DHow_Task_Scheduler_Works.rst7 While the task scheduler is not bound to any particular type of parallelism,
8 it was designed to work efficiently for fork-join parallelism with lots of forks.
9 This type of parallelism is typical for parallel algorithms such as `oneapi::tbb::parallel_for
12 Let's consider the mapping of fork-join parallelism on the task scheduler in more detail.
15 - Enable as many threads as possible, by creating enough job, to achieve actual parallelism
49 which causes temporary breadth-first execution that converts potential parallelism
50 into actual parallelism.
H A Dcancelling_nested_parallelism.rst7 Nested parallelism is canceled if the inner context is bound to the
17 As with all of the library's nested parallelism, you can control
H A DParallelizing_Flow_Graph.rst7 In addition to loop parallelism, the |full_name| library also supports graph parallelism.
12 Using graph parallelism, computations are represented by nodes and the
84 the runtime library spawns tasks to exploit the parallelism that is
94 The support for graph parallelism is contained within the namespace
H A DExceptions_and_Cancellation.rst17 there is oneTBB parallelism nested inside, the nested parallelism may
H A Duse_nested_algorithms.rst10 coarse-grained parallelism at the level of the graph, with finer grained
11 parallelism nested within.
H A DControlling_Chunking_os.rst86 the cost of reducing potential parallelism. The overhead as a fraction
114 parallelism. For example, if the grainsize is 1000 and the loop has 2000
119 parallel performance if there is other parallelism available higher up
H A DParallelizing_Simple_Loops_os.rst7 The simplest form of scalable parallelism is a loop of iterations that
H A DTask_Scheduler_Bypass.rst15 locality without adding significant parallelism. These problems can be avoided by using "Task Sched…
H A DThroughput_of_pipeline.rst11 experimentation. Too low a value limits parallelism; too high a value
H A DNon-Linear_Pipelines.rst25 parallelism is lost by forcing a linear order on the filters, but in
H A DCook_Until_Done_parallel_do.rst18 gain some parallelism.
H A DCancellation_and_Nested_Parallelism.rst7 The discussion so far was simplified by assuming non-nested parallelism
H A DTask-Based_Programming.rst13 - Matching parallelism to available resources
/oneTBB/doc/main/intro/
H A Dintro_os.rst10 Additionally, it fully supports nested parallelism, so you can build
28 The net result is that oneTBB enables you to specify parallelism far
H A Dtesting_approach.rst6 …our main types of errors/hazards you can encounter in the development of libraries for parallelism:
31 * The highest quality of the library is maintained even in such error-prone domains as parallelism.
H A DBenefits.rst23 - **oneTBB enables you to specify logical parallelism instead of
30 parallelism onto threads in a way that makes efficient use of
/oneTBB/doc/GSG/
H A Dintro.rst10 parallelism is carried out through threads, an operating system mechanism that allows the same or d…
25 oneTBB supports nested parallelism and load balancing. It means that you can use the library withou…
/oneTBB/doc/main/tbb_userguide/Migration_Guide/
H A DTask_Scheduler_Init.rst43 The default parallelism:
56 // Run the default parallelism
74 The limited parallelism:
97 // the default parallelism
/oneTBB/doc/main/tbb_userguide/design_patterns/
H A DAgglomeration.rst23 Many algorithms permit parallelism at a very fine grain, on the order
42 - The parallelism is for sake of performance and not required for
57 parallel overhead. Too large a block size may limit parallelism or
H A DLazy_Initialization.rst87 // but join parallelism inside functor
/oneTBB/
H A DREADME.md4 oneTBB is a flexible C++ library that simplifies the work of adding parallelism
12 * oneTBB enables you to specify logical parallelism instead of threads.
/oneTBB/test/conformance/
H A Dconformance_global_control.cpp67 void RunWorkersLimited(size_t parallelism) in RunWorkersLimited() argument
69 … oneapi::tbb::global_control s(oneapi::tbb::global_control::max_allowed_parallelism, parallelism); in RunWorkersLimited()
75 const std::size_t expected_threads = (utils::get_platform_max_threads()==1)? 1 : parallelism; in RunWorkersLimited()
/oneTBB/examples/task_group/sudoku/
H A DREADME.md4 It uses a straightforward state-space search algorithm that exhibits OR-parallelism. It can be opti…
/oneTBB/integration/linux/modulefiles/
H A Dtbb3246 module-whatis "Description: Flexible threading library for adding parallelism to complex applicatio…
H A Dtbb46 module-whatis "Description: Flexible threading library for adding parallelism to complex applicatio…

12