Solutions are available for a related question here. Go to have a check now!
It depends on how cloud servers are allocated.
If you perform a task that is not split, the local extraction will faster than using cloud extraction. Because only one cloud server will be allocated to the unsplit task and your machine works better than one cloud server.
If you perform a task that is split into sub-tasks in the cloud, then your task will split into 4 sub-tasks and 4 cloud servers will allocated to these sub-tasks (for Standard Edition). Thus executing tasks in the cloud will speed up the extraction and in this case have better performance than local extraction.
If you have any question, we'd happy to help.