Speed up MATLAB applications using Amazon EC2 and other cloud services
You can speed up your MATLAB programs or Simulink models by running them in a high-performance computing environment, such as those now offered by Amazon EC2 and other cloud computing services.
You use MATLAB and Parallel Computing Toolbox to enable your application to take advantage of desktop resources (multicore processors and GPUs). To solve larger problems you scale up using MATLAB Distributed Computing Server running on your computer cluster, or on Amazon EC2 or other cloud computing service.
This approach is a growing trend because users get instant access to the latest high-performance computers with large memory and GPUs with minimum upfront and maintenance costs.
Whether you run MATLAB Distributed Computing Server on the cloud or on your organization’s computer cluster, you have several licensing options. You can use it with on-demand (hourly) licensing as well as perpetual or annual licensing options.
- Launch computations on the cloud directly from MATLAB using Parallel Computing Toolbox
- License MATLAB Distributed Computing Server with on-demand (hourly) , perpetual, or annual options.
Note: Product use is billed by MathWorks, while use of the cloud computing infrastructure is billed by the cloud vendor (such as Amazon).
- Easily manage MATLAB clusters running on Amazon EC2 on the MathWorks Cloud Center site
- MATLAB (R2012a or later) and Parallel Computing Toolbox for creating your applications on the desktop
- Familiarity with Parallel Computing Toolbox concepts
- MATLAB Distributed Computing Server license for executing applications on the cloud
Note: MathWorks provides pre-built AMIs for use on Amazon EC2. For other clouds, work with your provider to set up MATLAB Distributed Computing Server.
Examples and How To
Amazon EC2 Resources
See also: MATLAB Distributed Computing Server, Parallel Computing Toolbox, Parallel computing, MATLAB multicore, MATLAB acceleration, Big data with MATLAB