Skip to content

Engineering Development Hub

Engineering Development Hub (EDH) is an open-source platform that brings cloud workstations, CAE simulation, and AI-supported engineering workflows together in one production-ready environment. Teams can access powerful Linux and Windows workstations, scale compute for demanding simulations, and use a built-in AI Assistant, all through a web browser.

Running entirely in your own AWS account, EDH gives your organization the flexibility to scale capacity with demand while keeping costs, data, and infrastructure under your control. Automated deployment gets teams up and running in hours, while built-in cost management, user administration, backups, SSO, persistent shared storage, and capacity planning make EDH practical to operate at scale.

Open-source and proven in real engineering environments since 2019, EDH gives you a transparent foundation you can own, adapt, and grow with your workloads.

New name, same solution

EDH was formerly known as Scale-Out Computing on AWS (SOCA) and was renamed in April 2026.

Why teams choose EDH

  • Up and running fast

    Deploy a complete environment in your own AWS account with a fully automated, one-step installation. No manual cluster build.

    Install your cluster

  • Powerful virtual desktops

    Launch Linux and Windows cloud desktops for your CAE and design applications, and connect from any browser in one click.

    Explore virtual desktops

  • HPC that scales on demand

    Submit jobs and let EDH pick and provision the right compute automatically, so users get capacity without touching infrastructure.

    Run HPC jobs

  • Built-in AI Assistant

    An assistant powered by Amazon Bedrock helps users understand their jobs, analyze files, and get answers in plain language.

    Meet the AI Assistant

  • Costs you can see and cap

    Track spend by user, team, project, or queue, set budgets, and automatically stop work when a budget is spent.

    Manage budgets and cost

  • Secure and managed

    Encryption everywhere, single sign-on, fine-grained access controls, automatic backups, and open-source code you fully own.

    Explore security

  • Storage for any workload

    From unlimited shared file systems to high-speed scratch, EDH supports the full range of AWS and external storage, so you can match cost and performance to each job.

    See storage options

  • Use your existing directory

    Connect to Active Directory or OpenLDAP, whether managed by EDH or one you already run, so users sign in with the accounts they already have.

    Set up your user directory

How to install EDH

Installation of your Engineering Development Hub cluster is fully automated and managed by AWS CloudFormation, so a complete environment stands up in your account without manual setup.

Once it is running, access your cluster in one click using DCV (Desktop Cloud Visualization)1 or through SSH.

Did you know?

  • You can run multiple Engineering Development Hub clusters on the same AWS account.
  • EDH tags its resources with a unique set of tags, making cost and resource tracking easy for AWS administrators.

Virtual desktops for every user

Launch persistent Linux and Windows virtual desktops to run your favorite CAE and engineering applications, from a lightweight session up to powerful GPU hardware. Users connect from any browser, so there is nothing to install locally.

Run large-scale compute without the complexity

Users submit a job and EDH takes care of finding, provisioning, and configuring the right hardware. Advanced users can fine-tune every detail, while everyone else can let EDH choose the most cost-effective option your administrators allow.

See a job submission example

Engineering Development Hub supports a set of parameters designed to simplify job submission on AWS. Advanced users can choose the compute, storage, and network configuration for a job, or simply skip these parameters and let EDH pick the most optimal hardware (defined by the HPC administrator).

# Advanced Configuration
user@host$ qsub -l instance_type=c5n.18xlarge \
    -l instance_ami=ami-123abcde
    -l nodes=2
    -l scratch_size=300
    -l efa_support=true
    -l spot_price=1.55 myscript.sh

# Basic Configuration
user@host$ qsub myscript.sh

EDH gives you flexibility in how capacity is provisioned and scheduled:

Do more with the AI Assistant

Engineering Development Hub includes an AI Assistant powered by Amazon Bedrock. Users ask questions in plain language, get help understanding their jobs and cluster, and analyze input and result files without leaving the web interface.

  • Chat directly with a large language model, talk to an agent that can act on your behalf, or use tools connected by your administrator.
  • Send a file straight from the File Explorer with Analyze with AI Assistant, then choose exactly which lines to send.
  • Every response shows its token usage and estimated cost, and each user has a daily usage limit so costs stay predictable.

Powered by Amazon Bedrock

Administrators choose which models are available, set the daily usage limit per user, and decide whether agents and tools are enabled. Learn more about the AI Assistant.

Keep costs under control

Cost visibility and guardrails are built in, so a shared environment never means runaway spend:

Analytics and business intelligence

Engineering Development Hub includes OpenSearch (formerly Elasticsearch) and automatically ingests job and host data in real time, giving you an accurate view of cluster activity.

Don't know where to start?

EDH includes dashboard examples if you are not yet familiar with OpenSearch or Kibana.

A simple web interface for everyone

Engineering Development Hub includes a web interface designed to make everyday tasks easy:

Automate with the HTTP REST API

Users can submit, retrieve, and delete jobs remotely through an HTTP REST API.

Secure your API access with scoped tokens

For scripts, automation, and CI/CD pipelines, use scoped API tokens instead of sharing your password. Each token can be limited to specific endpoints and methods, given an expiration, and revoked on its own if it is ever exposed. Administrators can also review all token activity in the token audit log.

An enterprise-ready foundation

Under the hood, EDH gives you the storage, security, and flexibility a shared engineering environment needs.

  • Bring your own images. Integrate custom AMIs with -l instance_ami=<ami_id> at job submission, with no limit on the number of AMIs. EDH supports heterogeneous environments, so concurrent jobs can run different operating systems on the same cluster.
  • Storage for any workload. From unlimited shared file systems to high-speed scratch, EDH supports the full range of AWS and external storage. See all storage options.
  • Connect to your existing directory. Use Active Directory or OpenLDAP, whether managed by EDH or one you already run. Set up your user directory.
  • Automatic backups. EDH backs up your data automatically, with no extra effort on your side.
  • Fully open-source and customizable. EDH is built entirely on AWS from CloudFormation templates, shell scripts, and Python, and the entire codebase is open-source on GitHub.
Running an AMI with a different OS than the scheduler

If your AMI uses a different operating system than your scheduler host, specify the OS manually so the right packages are installed for that node.

For example, if your Engineering Development Hub deployment uses Amazon Linux 2 but you want to submit a job on a personal RHEL7 AMI:

user@host$ qsub -l instance_ami=<ami_id> -l base_os=rhel7 myscript.sh

Connect to Active Directory or OpenLDAP

Engineering Development Hub works with the identity system you already use. You can run on Active Directory (through AWS Directory Service or an existing Active Directory) or on OpenLDAP (managed by EDH or your own existing OpenLDAP). Either way, users sign in with the accounts they already have, and you keep a single place to manage who has access. See all user directory options.

Storage for every kind of workload

Engineering Development Hub supports the full range of storage, so you can match cost and performance to each job. Options include Amazon EFS, Amazon EBS (including Provisioned IOPS), Amazon FSx for Lustre, FSx for NetApp ONTAP, FSx for OpenZFS, NVMe instance storage, and Amazon S3, as well as external on-premises NFS, NetApp, and Lustre file systems.

Storage comes in two flavors: shared storage that is mounted automatically on every node, and dedicated storage attached only to the nodes a specific job or desktop provisions. Explore all storage options.

Fair-share scheduling

Each user is given a score that varies based on:

  • Number of jobs in the queue
  • How long each job has been queued
  • The priority of each job
  • The type of instance requested

The job belonging to the user with the highest score starts next. Fair-share is configured per queue, so you can run one queue as FIFO and another as fair-share.

Automatic error handling

Engineering Development Hub runs several dry-run checks before provisioning capacity. If AWS cannot fulfill a request (for example, 5 instances are needed but only 3 are available in a placement group), EDH keeps trying for 30 minutes. If capacity is still unavailable after that, it automatically resets the request and tries a different availability zone. All of these errors are reported in the web interface to simplify troubleshooting.

Supported operating systems

Supported OS for Virtual Desktops, Compute Nodes, and Infrastructure Hosts:

Operating System Virtual Desktops Compute Nodes Infrastructure Host
Amazon Linux 2 ✅ ✅ ✅
Amazon Linux 2023 [1] ✅ ✅ ✅
Red Hat Enterprise Linux 7 ⚠️ ⚠️ ⚠️
Red Hat Enterprise Linux 8 [1] ✅ ✅ ✅
Red Hat Enterprise Linux 9 [1] ✅ ✅ ✅
CentOS 7 ⚠️ ⚠️ ⚠️
Rocky 8 [1] ✅ ✅ ✅
Rocky 9 [1] ✅ ✅ ✅
Ubuntu 22.04 [2] ✅ ✅ ✅
Ubuntu 24.04 [2] ✅ ✅ ✅
Windows Server 2019 ✅ ❌ ❌
Windows Server 2022 [2] ✅ ❌ ❌
Windows Server 2025 [2] ✅ ❌ ❌

[1] SOCA 2.7.5 and newer versions

[2] SOCA 25.3 and newer versions

  • ✅ Tested as of latest EDH version
  • ❌ Not supported
  • ⚠️ Operating System reached End Of Life. May or may not work without additional customization. No longer tested. See additional details

And more

Explore the tutorials, security, and analytics sections to learn more about everything Engineering Development Hub can do.

  1. DCV is a remote visualization technology that lets users securely connect to graphics-intensive 3D applications hosted on a remote high-performance server. ↩