In the post-build actions of your build job, mark the generated executables as artifacts, then you can use the Copy Artifact plugin to distribute the test executables to another test job (or more than one) that runs a Jenkins build slave on the test machine(s). As you've mentioned, you can configure a successful build to trigger the test jobs. Based on other answers, it looks like CUnit generates an XML report of the test output that Jenkins can parse, so in the test job's post-build actions, configure the location of the test results.
From a management perspective, it is easier if there is one test job because you don't have to figure out how to partition the executables and you can read the results in one report. But depending on your use case, it might make more sense to have separate test jobs if the tests require different environments or if it makes sense to partition the test results.