Concurrency & Parallel Runs

DeviceCloud supports parallel runs, also known as concurrency. This means multiple tests run simultaneously, greatly reducing the overall test suite execution time for a given CI run.

Limits are set per platform and runner type. In the console, under Settings → Execution, the Concurrency Limits card shows two values for each: minimum guaranteed concurrency (Concurrency Protection) and maximum concurrency (Surge Concurrency When Capacity Allows).

The plans come with these limits:

Plan	iOS default	iOS m4	Android cpu1 (default)	Android gpu1 (deprecated)	Android m1
Pro	up to 5	up to 5	up to 5	up to 2	up to 1
Max	up to 20, with 1 protected	up to 5	up to 20, with 1 protected	up to 2	up to 1

The gpu1 runner type is deprecated and will be removed on 9 November 2026, after which its runs use cpu1 and count towards the cpu1 limit. See Runner Type.

Cheaper plans typically have no Concurrency Protection, which means that during busy periods when runner capacity is stretched, you may need to wait for runner availability or experience less concurrency than your maximum.

Once you reach your maximum concurrency for a runner type, additional tests enter a queued state and sit in a backlog until you're under your concurrency limits. At that point, they automatically move to pending and are processed by the runners.

If you have a plan with Concurrency Protection, DeviceCloud will prioritise your jobs to ensure your minimum concurrency is met. On the Max plan, Concurrency Protection covers the standard runners only: iOS default and Android cpu1.

Some accounts have a shared pool instead, shown in the console as Standard — shared across iOS & Android. Standard iOS (default) and Android (cpu1) runs then draw on one combined limit rather than a limit each.

Enterprise Plans

If you need larger guaranteed concurrency or a bespoke concurrency setup and availability, please contact us at [e-mail] and we'd be happy to arrange an enterprise plan that meets your requirements.

Why limit concurrency at all?

DeviceCloud owns its own hardware, which is how we're able to offer such competitive prices. However, we have limited capacity in our runner pool. As such, we sometimes need to limit concurrency to ensure fair throughput for everyone. As more users subscribe, we gradually invest in and expand this pool to increase overall capacity.