Difference between the DPV Full, Split and Flat databases in CODE-1 Plus and Finalist

Products Affected:  CODE-1 Plus, Finalist
It is important for you to know which DPV file type best suits your configurations to ensure maximum product performance. The DPV file types are evaluated in detail below to help you identify the best DPV file type for your particular implementation.
DPV Full File 
The DPV Full File is the collection of the US delivery points from the United States Postal Service (USPS®) in one file. It consists of large binary hash files and requires a minimum of 650MB of RAM/Memory. In this implementation, the data for the entire United States is loaded at one time and stays in memory during processing.
DPV Split File 
The DPV Split File is created from the USPS’s DPV file of 100 binary hash files grouped by two-digit ZIP Code™.  For CODE-1 Plus, users will find that the DPV large memory module optimizes the use of the DPV Split File. The DPV large memory module requires up to 200MB of RAM/Memory. It is also possible to use the DPV small memory module which only loads data as needed.  
DPV Flat File 
The DPV Flat File is a flat database that contains one entry per delivery point and is sorted by five-digit ZIP Code. This file is in excess of 1.5GB and not dependent on RAM/Memory capacity. Input files should be sorted by ZIP Code. This is an optimal solution if you have less than 100MB of RAM/Memory available. Using Batch Local Shared Resources (BLSR) will provide the best performance.  Customers should use this file when extended memory is not available.
UPDATED:  June 18, 2019