The ParAccel Analytic Database (PADB) provides exactly what it says on the tin: a massively parallel, columnar database that has been optimised for analytics and ad hoc queries. For smaller companies without complex requirements it may be suitable for use as a general purpose data warehouse but it would not usually be appropriate where a major portion of the workload consists of queries that only address a single row or just a few rows, such as look-ups in a master data management environment.
Typical PADB implementations are in single figures or tens of terabytes. However, that may be misleading since the columnar nature of PADB means that the data is highly compressed. A more accurate guide to the scalability of ParAccel