Главная
Study mode:
on
1
Intro
2
Your Computer is Already Parallel
3
Drop-in Integration
4
Motivation: Parallelizing VLSI CAD Tools
5
Two Big Problems of Existing Tools
6
Example: An Iterative Optimizer
7
Need a New C++ Parallel Programming System
8
"Hello World" in Taskflow (Revisited)
9
"Hello World" in OpenMPO
10
Dynamic Tasking (Subflow)
11
Subflow can be Nested and Recurive
12
#3: Heterogeneous Tasking (cudaFlow)
13
Heterogeneous Tasking (cont'd)
14
Three Key Motivations
15
Conditional Tasking (Simple if-else)
16
Conditional Tasking (While/For Loop)
17
Conditional Tasking (Non-deterministic Loops)
18
Conditional Tasking (Switch)
19
Existing Frameworks on Control Flow?
20
Composable Tasking
21
Everything is Unified in Taskflow
22
Example: k-means Clustering
23
Submit Taskflow to Executor
24
Executor Scheduling Algorithm
25
Worker-level Scheduling
26
Application 1: VLSI Placement (cont'd)
27
Application 2: Machine Learning
Description:
Explore a comprehensive conference talk on Taskflow, a heterogeneous task graph programming system with control flow. Learn about modern C++ techniques for efficient parallel and heterogeneous computing, including CPU-GPU dependent tasks and conditional tasking interfaces. Discover how Taskflow addresses the challenge of balancing programmer productivity with performance scalability in complex computing applications. Examine real-world use cases, implementation strategies, and the work-stealing algorithm for dynamic resource balancing. Gain insights into the future of C++ programming for heterogeneous systems and how Taskflow can simplify parallel computing for developers.

Taskflow - A Heterogeneous Task Graph Programming System with Control Flow

CppNow
Add to list