Parallel Computing
R2026bParallel computing can speed a solver, especially if the objective function is time-consuming. You must have a Parallel Computing Toolbox™ license to use the built-in parallel computing capabilities of Global Optimization Toolbox.
Topics
- Parallel Computing in Global Optimization Toolbox
Use multiple processors to speed optimization.
- Parallel Computing with MultiStart and GlobalSearch
Distribute start points across processors with MultiStart, or use parallel gradient estimation with GlobalSearch.
- Parallel Computing with ga, gamultiobj, particleswarm, and paretosearch
Evaluate populations or poll points in parallel.
- Parallel Computing with patternsearch
Evaluate poll points and search functions in parallel with patternsearch.
- Parallel Computing with surrogateopt
Evaluate search points on workers asynchronously using parfeval.
- Minimizing an Expensive Optimization Problem Using Parallel Computing Toolbox
Explore the effectiveness of parallel computing using two solvers:
fminconandga.
