GridGain^® can be deployed as a distributed, horizontally scalable in-memory database (IMDB). It supports ACID transactions and can be used with SQL, key-value, compute, machine learning and other data processing APIs.

GridGain memory-centric storage allows the system to store data and indexes both in memory and on disk. However, the disk persistence is optional. When persistence is enabled, GridGain can store more data than can fit in the available memory and can act as a complete distributed SQL database.

When the whole data set and indexes fit in memory with persistence disabled, GridGain will function as an in-memory database supporting SQL plus all the existing APIs for memory-only use cases.

One of the distinguishing characteristics of GridGain SQL is the support for distributed SQL JOINs. This works in both, colocated and non-colocated fashions. When colocated, JOINs are executed on the local data available on each node without having to move large data sets across the network. Such a colocated approach provides the best scalability and performance in distributed clusters.

You should also check the GridGain distributed SQL and in-memory data grid features.