Parallel for-Loops (parfor) - MATLAB & Simulink (original) (raw)
Main Content
Use parallel processing by running parfor on workers in a parallel pool
Parallel Computing Toolbox™ supports interactive parallel computing and enables you to accelerate your workflow by running on multiple workers in a parallel pool. Use parfor to execute for-loop iterations in parallel on workers in a parallel pool. When you have profiled your code and identified slow for-loops, try parfor to increase your throughput. Develop parfor-loops on your desktop and scale up to a cluster without changing your code.
Functions
Functions and Classes
| parfor | Execute for-loop iterations in parallel on workers |
|---|---|
| parforOptions | Options set for parfor |
| parpool | Create parallel pool on cluster |
| parfeval | Run function on parallel pool worker |
| ticBytes | Start counting bytes transferred within parallel pool |
| tocBytes | Read how many bytes have been transferred since callingticBytes |
| send | Send data between clients and workers using a data queue |
| afterEach | Define a function to call when new data is received on a data queue |
| parallel.Pool | Parallel pool of workers |
| parallel.pool.DataQueue | Send and listen for data between client and workers |
Topics
Getting Started with parfor
- Decide When to Use parfor
Discover basic concepts of aparfor-loop, and decide when to use it. - Convert for-Loops Into parfor-Loops
Diagnose and fix commonparforproblems. - Ensure That parfor-Loop Iterations are Independent
Unlike afor-loop,parfor-loop Iterations have no guaranteed order. - Nested parfor and for-Loops and Other parfor Requirements
Learn how to deal with parallel nested loops. - Troubleshoot Variables in parfor-Loops
Discover variable requirements and classification inparfor-loops.
Running parfor-Loops
- Interactively Run Loops in Parallel Using parfor
Convert afor-loop into a scalableparfor-loop. - Improve parfor Performance
Create arrays inside or outsideparfor-loops to speed up code. - Run Code on Parallel Pools
Learn about starting and stopping parallel pools, pool size, and cluster selection. - Specify Your Parallel Preferences
Specify your preferences, and automatically create a parallel pool. - Use Objects and Handles in parfor-Loops
Discover how to use objects, handles, and sliced variables inparfor-loops. - Ensure Transparency in parfor-Loops or spmd Statements
All references to variables inparfor-loops must be visible in the body of the program. - Save Variables in parfor-Loops
This example shows how to use thesavefunction in aparfor-loop to save one or more variables to a file. (Since R2024a) - Scale Up parfor-Loops to Cluster and Cloud
Developparfor-loops on your desktop, and scale up to a cluster without changing your code. - Use parfor-Loops for Reduction Assignments
You can useparfor-loops to calculate cumulative values that are updated by each iteration. - Repeat Random Numbers in parfor-Loops
Control random number generation inparfor-loops by assigning a particular substream for each iteration. - Use parfor to Speed Up Monte-Carlo Code
This example shows how to speed up Monte-Carlo code by usingparfor-loops. - Interactively Import and Process Data in Parallel
This example shows how to import and process data simultaneously in an interactive parallel pool. (Since R2023b)
Deep Learning
- Use parfor to Train Multiple Deep Learning Networks (Deep Learning Toolbox)
This example shows how to use aparforloop to perform a parameter sweep on a training option.
Featured Examples
Use parfor to Speed Up Monte-Carlo Code
Speed up Monte-Carlo code by using parfor-loops. Monte-Carlo methods are found in many fields, including physics, mathematics, biology, and finance. Monte-Carlo methods involve executing a function many times with randomly distributed inputs. With Parallel Computing Toolbox, you can replace a for-loop with a parfor-loop to easily speed up code.
