Simply stated, solutions that use Relational Database Management Systems (RDBMS) offer higher performance, easier access to data, and more rapid modifications. This saves you both time and money over outdated systems with no Relational Database Manager available.
The IBM U² Database differentiates itself from the first normal form databases (e.g., Oracle, Sybase) in four areas: support for multiple programming environments, support for complex data models, greater performance for OLTP applications and ease of use. The IBM U² Database is able to offer a best-of -breed data management system tuned for mission-critical business applications that supports leading edge Object-oriented programming, state-of-the-art SQL and BASIC programming, and legacy COBOL programming.
The IBM U² Database is based on an extended relational data model. The SQL relational model is a subset of the IBM U²’s Database extended relational model. Extended relational structures are not restricted to SQL’s “first normal form”. The first normal form (1NF) model states that there may only be a single value in a cell of a table. There is always only a one-to-one relationship within a row of a table. This limits the data represented in the table to “atomic” data, similar to that stored in a spreadsheet. The extended relational model allows both one-to-one and one-to-many and one-to-many-to-many relationships within a single row of a table.
The most notable benefit of the extended relation is that it reduces the number of tables required in the database. Fewer tables results in databases that are much simpler to administer (for example, backups contain fewer tables), easier to program to (database programmers need to remember the structures of fewer tables), and much easier to query from. Having fewer tables also has several performance benefits. Fewer tables mean fewer joins. Joins are a very resource intensive activity, requiring complex optimization techniques, temporary tables to store intermediate query results, and lots of disk access. Having all the data necessary to complete a query, and to complete a transaction, in a single table eliminates the need to perform multi-table joins. The IBM U² Database does indeed support multi-table joins across extended relational tables. In these cases, the IBM U² Database employs the same optimization techniques as the large 1NF databases (use of indices, temporary tables and optimization routines).