just curious, would you mind elaborating the use case behind running tests sequentially across multiple different machines?
Usually I get the direct opposite of this question ("can we run tests in parallel across different machines?")
But in order to do so (off the top of my head) you would need an orchestration tool (some CI framework),
create multiple pipelines designated for the different sets of test cases you'd like to run on each of those separate machines (probably tag based pipelines to specify the remote execution location?), then trigger downstream pipelines in the sequence that you'd like.
As for the licenses... I want to say that you'd probably need floating licenses as is the case for CI executions, since the user executing the tests are usually triggered through an agent. And floating licenses make things more flexible in general.