In the context of computational tasks being divided among a certain number of "workers" (or cores), select the most accurate statements regarding the comparison of splitting a task among 10 workers as opposed to 5 workers. Choice 1 of 5:The improvement in efficiency when splitting tasks among more workers is an exponential function, regardless of the nature of the task. Choice 2 of 5:The efficiency of task completion invariably improves with an increased number of workers, given the parallel nature of computations. Choice 3 of 5:The effect of increasing the number of workers from 5 to 10 is subject to the specific nature of the task and the characteristics of the inter-task dependencies. If the task can be perfectly parallelized, then splitting among more workers may improve efficiency. However, for tasks with dependencies or sequential parts, adding more workers may not improve or could even decrease efficiency. Choice 4 of 5:The communication overhead always outweighs the benefits of increased parallelism when the task is divided among more workers. Choice 5 of 5:Amdahl's Law, which defines the maximum improvement gained by parallel processing, implies that splitting a task among more workers always improves performance