This is a package of tools for solving numerical and demanding processing of large volumes of data problems using multi-core processors, GPU and clusters. Parallel Computing Toolbox allows you to parallelize MATLAB applications thanks to its high-level cycles with parallel constructions, special types of arrays and quantitative parallel algorithms, without the need for CUDA or MPI programming.
This toolkit can be used in conjunction with Simulink to run parallel projects for simulating models in parallel. Parallel Computing Toolbox is also designed to run applications on multi-core desktop systems with up to 12 users, and everyone will have full access to all MATLAB tools. The solution allows you to get all the benefits of GPU GPUs. Parallel Computing Toolbox can scale to a cluster of several computers using the MATLAB Distributed Computing Server product.
◊Parallel loops (parfor) for running algorithms using task-parallel technology on composite processors
◊Support CUDA-compatible NVIDIA GPUs
◊Ability to run up to 12 users locally on desktop systems with a multiple processor
◊Support for work in clusters and computer networks (using MATLAB Distributed Computing Server)
◊Run online and batch process data for parallel applications
◊Distributed arrays and SPMD (Single Program Multiple Data) constructs for working with large amounts of data and data-parallel algorithms