function cudppSort in CUDPP


The two options, CUDPP_SORT_RADIX_GLOBAL and CUDPP_SORT_RADIX, can be selected as an sorting algorithm to cudppSort function.
What is different between them?

some descriptions about them as following

CUDPP_SORT_RADIX Radix sort within chunks, merge sort to merge chunks together
CUDPP_SORT_RADIX_GLOBAL Global radix sort across entire input, no merge

Global radix is running on global memory ?
Why does Global radix has no merge operation ?