Say we have one worker with 4 CPU cores. How does parallelism configured in Dataflow worker machines? Do we parallelize beyond # of cores?
Where would this type of infor
One worker thread is used per core, and each worker thread independently processes a chunk of the input space.