A write to Strongly-ordered memory is only considered complete when the transaction is acknowledged as complete by the memory or peripheral device.A write to Device memory is permitted to complete (from the processor's point of view) before it actually reaches the peripheral or memory device. Device memory is therefore faster for typical uses.
device memory might not be in order if two access are written/read to/from different slaves.
please check: cortex-a device memory access order limited in some range?