Cloud Center Release Notes
October 2025
MATLAB Parallel Server: Streamlined User Interface for Creating and Using Clusters
MATLAB® Parallel Server™ on Cloud Center now has a new user interface and underlying infrastructure. The operating system is updated to Ubuntu® 24.04. As a result, MATLAB Parallel Server users get improved performance, security, and user experience.
The configuration settings for the cloud machine configuration in Cloud Center are adapted from the following reference architecture available on GitHub®: MATLAB Parallel Server on Amazon Web Services.
Compatibility Considerations
If you have an existing R2023b to R2025a MATLAB Parallel Server cloud cluster that you created before October 22nd, 2025, your cloud cluster will no longer work after November 20th, 2025.
Follow these steps to upgrade your existing cloud cluster.
Log in to Cloud Center.
Create a new MATLAB Parallel Server cloud cluster.
Download any necessary files and data from your original cluster and move it to your new cluster.
Delete your original cloud cluster.
You can continue using your existing MATLAB Parallel Server cloud cluster until November 20th, 2025, after which date, any Ubuntu 20.04 cloud clusters will no longer be operational or discoverable in MATLAB or Cloud Center.
Functionality being removed or changed
| Functionality | Change | Use Instead |
|---|---|---|
| Automatically terminate cluster is now called Auto-Shutdown (hrs). The behavior of the Auto-Shutdown (hrs) operation is different from the Automatically terminate cluster feature. | When your cluster times out after the duration specified in your Auto-Shutdown (hrs) setting, your cluster machines stop, but other cluster resources continue to run and incur costs. These resources maintain your existing cluster access rules and allow you to restart the cluster quickly. | To avoid costs, stop these resources and save a data backup of the cluster in your cloud account by choosing Stop All Resources in the warning message. |
The Global Cluster Access setting has been removed. | With the Global Cluster Access setting, you could specify the list of IP ranges for the computers that can access your cloud cluster. This list of IP ranges applied to all your running clusters. | You can use the Default IP Address setting to specify a list of IP addresses that can access the cluster. After you start the cluster, you can configure the IP address that can access the cluster in the Access Ports and IP section. For details, see Manage Cluster Access. You need to configure the settings for each cluster individually. |
The custom cluster access rules that you add do not persist when you Stop a cluster and restart it. | Previously, custom cluster access rules persisted when you stopped a cluster and started it again. | You can customize the cluster access rules again after you stop and restart a cluster. For details on how to do this, see Manage Cluster Access. You can also use the Default IP Address setting to specify a list of IP addresses that can access the cluster. This list of IP addresses persists when you stop and restart the cluster. |
The Use a Dedicated headnode configuration setting has been removed. Instead, a dedicated headnode is used by default for all clusters. | You could previously select whether you wanted a dedicated headnode with your cluster. Now, a dedicated headnode is attached to your cluster by default. All management services (for example, the job manager or the file server) run on the headnode and not the worker nodes. This ensures that running computations that use many system resources, such as memory, processor, network, or local storage, does not negatively impact the performance of the job manager. | - |
| The Local Machine Storage configuration settings have been removed. | You can no longer enter the ID of an EBS snapshot when creating a cluster to access all data in the original snapshot. You can also no longer create a local data volume on each worker machine when you create a cluster. | - |
| Persisted Storage is now called Shared Storage Size (GB). To specify a shared storage size, you must select the checkbox Enable shared storage. | Previously, if you did not specify Persisted Storage, your MATLAB Job Scheduler database did not persist when you stopped and restarted the cluster. Now, regardless of whether you select Enable shared storage, your MATLAB Job Scheduler always persists when you stop and restart the cluster. | - |
The Operating System Image (AMI) configuration setting has been removed. | Previously, you could use a custom Amazon® Machine Image (AMI) when creating a cloud cluster. This feature is no longer available. To learn more about the AMI being used in the cloud clusters, see the GitHub repository: MATLAB Parallel Server on Amazon Web Services. | - |
| The Upload S3 Data configuration setting has been removed. | You can no longer upload Amazon S3™ data while creating your cluster. | You can upload Amazon S3 data to your cluster after you have created it. See Transfer Data to Amazon S3 Buckets and Access Data Using MATLAB for more details. |
You can no longer download the SSH private key for the username
clouduser by default. | By default, you cannot use the SSH private key for the username
clouduser to transfer data into your cluster
using standard utilities such as SCP, SFTP, FileZilla, etc. | To SSH into the cluster using the username
clouduser, your cluster administrator must
set up the user profile for you. For details, see Download SSH Key Identity File. You can then use this
profile to transfer data to the cluster. For details, see Transfer Data with Standard Utilities. |
| The Highest option for the Cluster log level setting has been removed. | Previously, you could set Highest for Cluster log level in Cloud Center. | - |
October 2025
Support for MATLAB R2025b
You can run MATLAB R2025b on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB
Support for MATLAB R2025a
You can run MATLAB R2025a on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB
May 2025
Support for MATLAB Parallel Server with R2025a
You can run clusters with R2025a on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
September 2024
Support for MATLAB R2024b
You can run MATLAB R2024b on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB.
Support for MATLAB Parallel Server with R2024b
You can run clusters with R2024b on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
July 2024
Functionality being removed or changed
Support for releases MATLAB R2022a and MATLAB R2021b will be removed in September 2024.
To check supported releases, see Supported Releases for MATLAB.
April 2024
Support for MATLAB R2024a
You can run MATLAB R2024a on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB.
Support for MATLAB Parallel Server with R2024a
You can run clusters with R2024a on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
February 2024
Functionality being removed or changed
AWS® instance types cc2.8xlarge, cg1.4xlarge, g2.2xlarge, and g2.8xlarge are no longer available in most AWS regions, and their support will be removed from Cloud Center. If you are using any of these instance types in your Cloud Center cluster, configure your cluster using other supported instance types.
For a list of currently supported instance types, see Choose Supported EC2 Instance Machine Types.
September 2023
Support for MATLAB R2023b
You can run MATLAB R2023b on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB.
Support for MATLAB Parallel Server with R2023b
You can run clusters with R2023b on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
April 2023
Support for MATLAB R2023a
You can run MATLAB R2023a on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB.
March 2023
Support for MATLAB Parallel Server with R2023a
You can run clusters with R2023a on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
Support for new instance classes
Support for five new General Purpose instance classes: m6a, m6i, m6id, m6in, and m6idn.
Support for five new Compute Optimized instance classes: c6a, c6i, c6id, and c6in.
Support for a new Accelerated Computing (GPUs) instance classes: g5.
Support for five new Memory Optimized classes: r6a, r6i, r6id, r6in, and r6idn.
For more details on newly added compute optimized instances and regional availability, see Choose Supported EC2 Instance Machine Types.
October 2022
Support for MATLAB R2022b
You can run MATLAB R2022b on Cloud Center instances.
To check supported releases, see Supported Releases for MATLAB.
September 2022
Support for MATLAB Parallel Server with R2022b
You can run clusters with R2022b on Cloud Center instances.
To check supported releases, see Supported Releases for Clusters.
April 2022
Run MATLAB in AWS Using Cloud Center
Now you can use Cloud Center to start MATLAB in Amazon Web Services (AWS). You can start a single machine with MATLAB installed that you can access from a web browser or a remote desktop application. To get started, see Link Cloud Account to Cloud Center and Start MATLAB on Amazon Web Services (AWS) Using Cloud Center.
You can continue to use Cloud Center to start and manage MATLAB Parallel Server clusters that you can access from any MATLAB. To check supported releases, see Supported Releases for Clusters.
September 2021
Support for MATLAB Release R2021b
You can run MATLAB R2021b on Cloud Center instances.
May 2021
Support for new instance classes
Support for six new General Purpose instance classes: m5a, m5ad, m5d, m5dn, m5n and m5zn.
Support for five new Compute Optimized instance classes: c5, c5a, c5ad, c5d, and c5n.
Support for two new Accelerated Computing (GPUs) instance classes: g4dn and p3dn.
Support for eight new Memory Optimized classes: r5, r5a, r5ad, r5b, r5d, r5dn, r5n, and x1e.
Note
c5d.xlarge is the new default headnode instance type and m5.8xlarge is the new default worker instance for clusters in a VPC.
For more details on newly added compute optimized instances and regional availability, see Choose Supported EC2 Instance Machine Types.
March 2021
Support for MATLAB Release R2021a
You can run MATLAB R2021a on Cloud Center instances.
MATLAB Worker Amazon Machine Images (AMI) now support Ubuntu 20.04 LTS
All worker AMIs now include installations of Ubuntu 20.04.
MATLAB worker Amazon Machine Images (AMI) now support CUDA 11.0 and gcc/g++ 9.3
All worker AMIs now include installations of CUDA Toolkit 11.0 and gcc/g++ 9.3. You can generate CUDA kernel objects from CU code or compile CUDA compatible source code, libraries, and executables using GPU Coder™ on GPU-enabled EC2 instances with supported GPU devices. For information on supported devices, see GPU Computing Requirements (Parallel Computing Toolbox). You can use MathWorks® products (such as MATLAB Coder™ or HDL Coder™) that require a gcc/g++ 9.3 or earlier compiler on MATLAB worker AMIs. For information on product requirements, see System Requirements and Supported Compilers.
September 2020
Support for MATLAB Release R2020b including one additional pretrained convolutional neural networks (CNN) for deep learning
You can run MATLAB R2020b on Cloud Center instances, including support for one additional pretrained convolutional neural network (CNN) model: EfficientNet-b0. You can use this pretrained model for classification and transfer learning. You can access the model using the function
efficientnetb0(Deep Learning Toolbox). For details, see Parallel and Cloud (Deep Learning Toolbox) and Pretrained Deep Neural Networks (Deep Learning Toolbox).MATLAB worker Amazon Machine Images (AMI) now support CUDA 10.2 and gcc/g++ 6.5
All worker AMIs now include installations of CUDA Toolkit 10.2 and gcc/g++ 6.5. You can generate CUDA kernel objects from CU code or compile CUDA compatible source code, libraries, and executables using GPU Coder on GPU-enabled EC2 instances with supported GPU devices. For information on supported devices, see GPU Computing Requirements (Parallel Computing Toolbox). You can use MathWorks products (such as MATLAB Coder or HDL Coder) that require a gcc/g++ 6.5 or earlier compiler on MATLAB worker AMIs. For information on product requirements, see System Requirements and Supported Compilers.
March 2020
Functionality being removed or changed
| Functionality | Result | Use Instead | Compatibility Considerations |
|---|---|---|---|
A Shareable Cluster no longer requires the “Shared With” white list (applies to all releases, except R2019b which still requires the “Shared With” white list). | Anyone with the cluster profile can access the shared cluster (except in R2019b, which still requires the "Shared With" white list). | To restrict access with a "Shared With" white list, use R2019b only. | An existing "Shared With" white list has no effect on cluster access in any version except R2019b and should not be used in any other release with the expectation that this list would restrict user access to a cluster. |
| You can no longer export personal cluster profiles. | For Personal clusters, the export cluster profile option is no longer available. Personal clusters were introduced in R2019b. You can still export cluster profiles for any R2019b Shareable cluster and for any R2018a - R2019a cluster. | Cluster profiles for running personal clusters can be added to MATLAB via the Discover Clusters... option under the Parallel menu in MATLAB. Cluster profiles are automatically added to MATLAB by creating a new personal clusters via the Create Cloud Cluster option on the Cluster Profile Manager in MATLAB. The Cluster Profile Manager is accessed via the Create and Manage Clusters... option under the Parallel menu in MATLAB | None |
October 2019
Cluster Shared State
The cluster attribute Shared State enables Cloud Center to authorize users to submit jobs to or interact with a cluster from MATLAB. The Shared State may be Personal Cluster (accessible only by you) or Shareable Cluster (accessible to people you have explicitly given access to via a white list). The default Shared State attribute of a cluster is Personal Cluster. This authorization control is unrelated to accessing the cluster via SSH.
For Shareable Clusters, the person creating the cluster must expressly add authorized users to the Shared With field.
Note
Cloud Center clusters created for MATLAB releases prior to R2019b did not have clusters with the Shared State attribute. Those clusters are all implicitly shareable clusters, as any user who imports a cluster profile is authorized to submit jobs to or interact with the cluster from MATLAB. The inbound firewall rules for these clusters are managed only by Global Cluster Access rules.
Auto-Manage Cluster Access
The cluster attribute Auto-Manage Cluster Access allows Cloud Center to manage a cluster’s inbound firewall rules on a cluster-by-cluster basis. Auto-Manage Cluster Access is enabled by default for Personal Clusters.
Note
Clusters created prior to R2019b are Shareable Clusters, whose inbound firewall rules are managed only by Global Cluster Access rules.
For more information, see Manage Cluster Access.
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2017b has been removed. For more information on the migration policy, see Requirements for Using Cloud Center. Errors MATLAB versions R2019b, R2019a, R2018b, or R2018a As newer versions of MATLAB become supported, the support for older versions will be removed in future releases.
June 2019
Support for MATLAB Release R2019a Update 3 including three additional pretrained convolutional neural networks (CNN) for deep learning
You can run MATLAB R2019a Update 3 on Cloud Center instances, including support for three additional pre-trained convolutional neural network (CNN) models: NASNet-Large, NASNet Mobile, and ShuffleNet. You can use these pretrained models for classification and transfer learning. You can access the models using the functions
nasnetlarge(Deep Learning Toolbox),nasnetmobile(Deep Learning Toolbox), andshufflenet(Deep Learning Toolbox). For details, see Parallel and Cloud (Deep Learning Toolbox) and Pretrained Deep Neural Networks (Deep Learning Toolbox).
MathWorks sign-on includes Cloud Center sign-on
Signing in to MathWorks simultaneously signs you in to Cloud Center using your MathWorks Account.
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2017a has been removed. For more information on the migration policy, see Requirements for Using Cloud Center. Errors MATLAB versions R2019a, R2018b, R2018a, or R2017b As newer versions of MATLAB become supported, the support for older versions will be removed in future releases.
March 2019
Support for MATLAB Release R2019a including pretrained convolutional neural networks (CNN) for deep learning
You can run MATLAB Release R2019a on Cloud Center instances, including support for pretrained convolutional neural network (CNN) models. These networks are now available on all instances for MATLAB Release R2019a: AlexNet, DenseNet-201, GoogLeNet (trained using ImageNet and Places365 data sets), Inception-ResNet-v2, Inception-v3, MobileNet-v2, ResNet-18, ResNet-50, Resnet-101, SqueezeNet, VGG-16, VGG-19, and Xception. You can access the models using the functions
alexnet(Deep Learning Toolbox),densenet201(Deep Learning Toolbox),googlenet(Deep Learning Toolbox),inceptionresnetv2(Deep Learning Toolbox),inceptionv3(Deep Learning Toolbox),mobilenetv2(Deep Learning Toolbox),resnet18(Deep Learning Toolbox),resnet50(Deep Learning Toolbox),resnet101(Deep Learning Toolbox),squeezenet(Deep Learning Toolbox),vgg16(Deep Learning Toolbox),vgg19(Deep Learning Toolbox), andxception(Deep Learning Toolbox). You can use these pretrained models for classification and transfer learning. For details, see Parallel and Cloud (Deep Learning Toolbox), and Pretrained Deep Neural Networks (Deep Learning Toolbox).
Automatic Cluster Resizing: Resize Cloud Center clusters on Amazon based on usage
You can create cloud clusters using MATLAB Release R2019a that resize automatically based on usage. These clusters grow or shrink to allocate the optimal number of workers for your submitted tasks. For more information, see Resize Clusters Automatically.
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2017a will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2019a, R2018b, R2018a, or R2017b. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
January 2019
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016b has been removed. For more information on the migration policy, see Requirements for Using Cloud Center. Errors MATLAB versions R2018b, R2018a, R2017b, or R2017a. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
December 2018
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for the Amazon EC2® Classic network type has been removed. Errors Set up a new cluster using the Amazon EC2 Virtual Private Cloud (VPC) network type. Amazon Web Services withdrew support for the EC2 Classic network for new accounts in 2013. To continue using MATLAB with cloud resources, see Configure AWS VPC for Cloud Center. Cloud Center support for running MATLAB version R2016b will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2018b, R2018a, R2017b, or R2017a. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
October 2018
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for the Amazon EC2 Classic network type will be removed in a future release. Warns Set up a new cluster using the Amazon EC2 Virtual Private Cloud (VPC) network type. Amazon Web Services withdrew support for the EC2 Classic network for new accounts in 2013. To continue using MATLAB with cloud resources, see Configure AWS VPC for Cloud Center. Cloud Center support for running MATLAB version R2016b will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2018b, R2018a, R2017b, or R2017a. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
September 2018
Support for MATLAB Release R2018b including pretrained convolutional neural networks (CNN) for deep learning
You can run MATLAB Release R2018b on Cloud Center instances, including support for pretrained convolutional neural network (CNN) models. These networks are now available on all instances for MATLAB Release R2018b: AlexNet, DenseNet-201, GoogLeNet, Inception-ResNet-v2, Inception-v3, ResNet-18, ResNet-50, Resnet-101, SqueezeNet, VGG-16, and VGG-19. You can access the models using the functions
alexnet(Deep Learning Toolbox),densenet201(Deep Learning Toolbox),googlenet(Deep Learning Toolbox),inceptionresnetv2(Deep Learning Toolbox),inceptionv3(Deep Learning Toolbox),resnet18(Deep Learning Toolbox),resnet50(Deep Learning Toolbox),resnet101(Deep Learning Toolbox),squeezenet(Deep Learning Toolbox),vgg16(Deep Learning Toolbox), andvgg19(Deep Learning Toolbox). You can use these pretrained models for classification and transfer learning. For details, see Parallel and Cloud (Deep Learning Toolbox), and Pretrained Deep Neural Networks (Deep Learning Toolbox).
MATLAB worker Amazon Machine Images (AMI) now support CUDA 9.0
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016b will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2018b, R2018a, R2017b, or R2017a. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
August 2018
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016a has been removed. For more information on the migration policy, see Requirements for Using Cloud Center. Errors MATLAB versions R2018a, R2017b, R2017a, or R2016b. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
April 2018
Create clusters in dedicated headnode mode
Starting this release, you can create clusters in dedicated headnode mode. When enabled, the headnode instance exclusively runs management services, such as the job manager, and does not host any MATLAB workers. This cluster architecture improves performance.
Increased maximum number of workers to 1024
You can now create up to
1024worker machines in VPC networks for MATLAB R2018a onwards.
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016a will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2018a, R2017b, R2017a, or R2016b. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
March 2018
Support for MATLAB Release R2018a including pretrained convolutional neural networks (CNN) for deep learning
You can run MATLAB Release R2018a on Cloud Center instances, including support for pretrained convolutional neural network (CNN) models. These networks are now available on all instances for MATLAB Release R2018a: AlexNet, VGG-16, VGG-19, GoogLeNet, ResNet-50, and ResNet-101. You can access the models using the functions
alexnet(Deep Learning Toolbox),vgg16(Deep Learning Toolbox),vgg19(Deep Learning Toolbox),googlenet(Deep Learning Toolbox),resnet50(Deep Learning Toolbox), andresnet101(Deep Learning Toolbox). You can use these pretrained models for classification and transfer learning. For details, see Parallel and Cloud (Deep Learning Toolbox), and Pretrained Deep Neural Networks (Deep Learning Toolbox).
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016a will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns MATLAB versions R2018a, R2017b, R2017a, or R2016b. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
February 2018
Functionality being removed or changed
Functionality Result Use Instead Compatibility Considerations Cloud Center support for running MATLAB version R2016a will be removed in a future release. For more information on the migration policy, see Requirements for Using Cloud Center. Warns Migrate to MATLAB versions R2017b, R2017a, or R2016b. As newer versions of MATLAB are supported, the support for older versions will be removed in future releases.
