GitHub partners with Arm to revolutionize Internet of Things software development with GitHub Actions

Arm Toolchain Lands in GitHub Actions Runners

Arm’s own survey data suggests the shift to remote work is reshaping how IoT design teams operate. More than 70 percent of respondents said the distributed workforce is a key factor in future device design, and 93 percent expect remote development models to persist. That trend puts pressure on toolchains that historically required local hardware and manually configured environments.

GitHub and Arm have teamed up to address that friction by making Arm’s development toolchain—including the cross-compiler and Arm Virtual Hardware—available natively inside GitHub Actions cloud-hosted runners. The goal is to give IoT teams a CI workflow that scales horizontally without requiring them to build or manage their own board farms, and without the licensing overhead that typically comes with proprietary tooling.

From Desktop to Cloud Without Hardware Dependencies

IoT developers—particularly those working on safety-critical systems in industrial automation, medical, automotive, railway, or avionics—rely on a complex stack: compiler, debugger, and simulation tooling. Historically, that stack lived on physical desktops or customer-managed CI environments. Arm Virtual Hardware removes the dependency on physical boards by running as a cloud application, simulating memory and peripherals so tests can execute without a hardware lab.

What held teams back was the cost of adoption: setting up that software stack inside a new CI/CD platform takes time, and proprietary tools traditionally required licensing tied to privately managed infrastructure. With the Arm tools preconfigured on the GitHub Actions runner, developers skip the setup phase and can start validating code changes in minutes. The runner includes all necessary Arm development tools, so teams can trigger pipelines directly from a pull request and get results without standing up any special hardware. In a demo using the Arm TS Keyword project—an embedded workload with roughly 500 source files—a code change committed to a branch kicked off a full pipeline: tool installation, Arm compilation, test execution on Arm Virtual Hardware, and pass/fail reporting back on the pull request. All of it ran on GitHub’s platform, with no customer-managed infrastructure.

Billing Moves from Licenses to Workloads

A key shift in this partnership is the commercial model. Instead of estimating how many licenses are needed to cover peak workload and paying for idle capacity, organizations pay based on the workloads actually run. Horizontal scalability is handled by the platform: teams can spin up as many or as few runners as a particular validation cycle requires. Once the jobs finish, results are presented to the developer, and the infrastructure disappears.

For a demo video and step-by-step instructions on running Arm Virtual Hardware inside GitHub Actions, see the Arm-software/AVH-GetStarted repository on GitHub.