MATLAB in Amazon Web Services

Work with data in the cloud, scale to high-performance clusters, and deploy your applications

Analyze Data in the Cloud with MATLAB and Simulink 

Run MATLAB® and Simulink® directly on EC2 instances in the Amazon Web Services (AWS) environment, allowing you to co-locate processing with your data. This lets you efficiently access and analyze your data using AWS storage services. Run analyses and simulations even faster by taking advantage of on-demand, high-performance CPU and GPU hardware. Get up and running on AWS with one of the following options: A MATLAB reference architecture, a deep learning container for NVIDIA® GPU Cloud (NGC), or AppStream 2.0.

This reference architecture lets you quickly create, configure, and use AWS infrastructure with MATLAB pre-installed. It incorporates best practices for creating a full MATLAB desktop experience. This includes connecting to AWS from your local desktop using Remote Desktop Protocol (RDP). It sets up a single instance that mounts a snapshot containing MATLAB, a private VPC with an internet gateway, a private subnet, and a security group that opens the appropriate ports for SSH and RDP access. Learn how by watching this instructional video (5:18).

The MATLAB deep learning container speeds up your deep learning applications by taking advantage of high-performance NVIDIA GPUs. It provides algorithms, pretrained models, and apps to create, train, visualize, and optimize deep neural networks. You can also access tools for image and signal processing, text analytics, and automatically generating C and CUDA code for deployment on NVIDIA GPUs in data centers and embedded systems. Watch a video that explains how to how to start the container on an AWS instance (15:25).

AppStream 2.0 is a fully managed application streaming service that can stream MATLAB from the cloud to your computers. It integrates with your network and cloud storage, eliminating the need to upload or download data. This getting started guide shows how to install and configure MATLAB on AppStream 2.0, perform foundational administrative tasks, and provision a private VPC to host your AppStream 2.0 resources. Because AppStream 2.0 is fully managed, you can scale to different regions without provisioning or operating any infrastructure.

Scale Your Applications and Simulations with MATLAB Distributed Computing Server

Scale computationally intensive and parallelizable MATLAB programs and Simulink models on clusters in AWS using MATLAB Distributed Computing Server. Go beyond using just a single computer, and take advantage of CPUs and GPUs across multiple instances without needing to recode. Reference architectures developed for AWS let you quickly create, configure, and use high-performance clusters for parallel computing.

This reference architecture incorporates best practices for running a full MATLAB Distributed Computing Server environment on AWS. This includes creating a profile to use when connecting from a MATLAB and Parallel Computing Toolbox desktop session. It sets up a cluster of computers using the MATLAB job scheduler, and ensures that all eligible products are installed. This means you can upload, store, install, and configure the software in less time. Available in limited AWS regions. Alternatively, you can prepare your own images or single AMI with software pre-installed.

For a ready-to-use option, learn more about Cloud Center.

Deploy Analytics with MATLAB Production Server

Incorporate MATLAB analytics into enterprise, web, and mobile applications running on AWS infrastructure with MATLAB Production Server™. The elasticity of AWS infrastructure combined with MATLAB Production Server enables your applications to support many users simultaneously. Reference architectures developed for AWS let you quickly create, configure, and deploy a MATLAB Production Server environment.

This reference architecture provides best practices to help you configure and deploy MATLAB Production Server in AWS. MathWorks provides templates that use preconfigured AMIs to create the necessary MATLAB Production Server and license server instances. This means you can install and configure the software and connect the servers to each other in less time. Available in limited AWS regions. If you prefer full control over your infrastructure, you can provision your own software onto your existing instances.

Operationalize IoT analytics using AWS IoT and MATLAB Production Server, or scale up IoT proof-of-concept systems that you prototyped in ThingSpeak