<h2><a class="anchor" id="GeneralQuestion8"></a>

Q8: What is the overhead of taskflow?</h2>

<p>Creating a taskflow has certain overhead. For example, creating a task and a dependency takes about 61 and 14 nanoseconds in our system (Intel 4-core CPU at 2.00GHz). The time is amortized over 1M operations, since we have implemented an object pool to recycle tasks for minimal overhead.</p>

<h2><a class="anchor" id="GeneralQuestion9"></a>

Q9: How does it compare to existing task programming systems?</h2>

<p>There is a large amount of work on programming systems (e.g., StarPU, Intel TBB, OpenMP, PaRSEC, Kokkos, HPX) in the interest of simplifying the programming complexity of parallel and heterogeneous computing. Each of these systems has its own pros and cons and deserves a reason to exist. However, they do have some problems, particularly from the standpoint of ease of use, static control flow, and scheduling efficiency. Cpp-Taskflow addresses these limitations through a simple, expressive, and transparent graph programming model.</p>

<h2><a class="anchor" id="GeneralQuestion10"></a>

Q10: Do you try to simplify the GPU kernel programming?</h2>

<p>No, we do not develop new programming models to simplify the kernel programming. The rationale is simple: Writing efficient kernels requires domain-specific knowledge and developers often require direct access to the native GPU programming interface. High-level kernel programming models or abstractions all come with restricted applicability. Despite non-trivial kernel programming, we believe what makes heterogeneous computing difficult is the surrounding tasks. A mistake made in task scheduling can outweigh all speed-up benefits from a highly optimized kernel. Therefore, Cpp-Taskflow focuses on heterogeneous tasking that affects the overall system performance to a large extent. </p><hr/>

<h1><a class="anchor" id="ProgrammingQuestions"></a>

Programming Questions</h1>

<h2><a class="anchor" id="ProgrammingQuestions1"></a>

<div class="contents">

<div class="textblock"><h1><a class="anchor" id="ModernCppParallelTaskProgrammingLibrary"></a>

Modern C++ Parallel Task Programming</h1>

<p>Cpp-Taskflow helps you quickly write parallel task programs with high performance scalability and simultaneous high productivity. It is faster, more expressive, fewer lines of code, and easier for drop-in integration than many of existing task programming frameworks.</p>

<div class="image">

<img src="performance.png" alt="performance.png" width="95%"/>

<p>Cpp-Taskflow is committed to support both academic and industry research projects, making it reliable and cost-effective for long-term and large-scale developments. Our users say:</p>

[ "Q5: What is the weird hex value, like 0x7fc39d402ab0, in the dumped graph?", "FAQ.html#GeneralQuestion5", null ],

[ "Q6: Does Cpp-Taskflow have backward compatibility with C++03/98/11?", "FAQ.html#GeneralQuestion6", null ],

[ "Q7: How does Cpp-Taskflow schedule tasks?", "FAQ.html#GeneralQuestion7", null ],

[ "Q8: What is the overhead of taskflow?", "FAQ.html#GeneralQuestion8", null ],

[ "Q9: How does it compare to existing task programming systems?", "FAQ.html#GeneralQuestion9", null ],

[ "Q10: Do you try to simplify the GPU kernel programming?", "FAQ.html#GeneralQuestion10", null ]

] ],

[ "Programming Questions", "FAQ.html#ProgrammingQuestions", [

[ "Q1: What is the difference between Cpp-Taskflow threads and workers?", "FAQ.html#ProgrammingQuestions1", null ],

var NAVTREEINDEX =

var SYNCONMSG = 'click to disable panel synchronisation';

"Examples.html":[4],

"FAQ.html":[5],

"FAQ.html#GeneralQuestion1":[5,0,0],

"FAQ.html#GeneralQuestion10":[5,0,9],

"FAQ.html#GeneralQuestion2":[5,0,1],

"FAQ.html#GeneralQuestion3":[5,0,2],

"FAQ.html#GeneralQuestion4":[5,0,3],

"FAQ.html#GeneralQuestion5":[5,0,4],

"FAQ.html#GeneralQuestion6":[5,0,5],

"FAQ.html#GeneralQuestion7":[5,0,6],

"FAQ.html#GeneralQuestion8":[5,0,7],

"FAQ.html#GeneralQuestion9":[5,0,8],

"FAQ.html#GeneralQuestions":[5,0],

"FAQ.html#ProgrammingQuestions":[5,1],

"FAQ.html#ProgrammingQuestions1":[5,1,0],

"graphtraversal.html#GraphTraversalProblemFormulation":[4,4,0],

"graphtraversal.html#GraphTraversalStaticTraversal":[4,4,2],

"hierarchy.html":[7,2],

"index.html":[0],

"index.html":[],

"index.html#ASimpleFirstProgram":[0,6],

"index.html#CompileAndRunYourFirstProgram":[0,7],

"index.html#ComposableTasking":[0,3],

"index.html#SupportedCompilers":[0,8],

"install.html":[1],

"install.html#BAIBuildExamplesTestsAndBenchmarks":[1,3],

"install.html#BAICompileAndLinkFlag":[1,1]

var NAVTREEINDEX1 =

"install.html#BAISettingUpCpp-Taskflow":[1,0],

"install.html#BAISupportedCompilers":[1,2],

"kmeans.html":[4,2],

"kmeans.html#BuiltInPredicate":[4,2,3],

"kmeans.html#KMeansBenchmarking":[4,2,4],

The time is amortized over 1M operations, since we have implemented an object pool

to recycle tasks for minimal overhead.

@subsection GeneralQuestion9 Q9: How does it compare to existing task programming systems?

There is a large amount of work on programming systems (e.g., StarPU, Intel TBB, OpenMP, PaRSEC, Kokkos, HPX) in the interest of simplifying the

programming complexity of parallel and heterogeneous computing.

Each of these systems has its own pros and cons and deserves a reason to exist.

However, they do have some problems, particularly from the standpoint of ease of use,

static control flow, and scheduling efficiency.

Cpp-Taskflow addresses these limitations through a simple, expressive, and transparent

graph programming model.

@subsection GeneralQuestion10 Q10: Do you try to simplify the GPU kernel programming?

No, we do not develop new programming models to simplify the kernel programming.

The rationale is simple: Writing efficient kernels requires domain-specific knowledge

and developers often require direct access to the native GPU programming interface.

High-level kernel programming models or abstractions all come with restricted applicability.

Despite non-trivial kernel programming, we believe what makes heterogeneous computing difficult

is the surrounding tasks.

A mistake made in task scheduling can outweigh all speed-up benefits from a highly optimized kernel.

Therefore, Cpp-Taskflow focuses on heterogeneous tasking

that affects the overall system performance to a large extent.

----

Cpp-Taskflow helps you quickly write parallel task programs with high performance scalability

and simultaneous high productivity.

It is faster, more expressive, fewer lines of code, and easier for drop-in integration

than many of existing task programming frameworks.

@image html images/performance.png width=95%

Cpp-Taskflow is committed to support both academic and industry research projects,

making it reliable and cost-effective for long-term and large-scale developments.