| 
| 
Why Select Lava Distributed SQL?
Raw Performance
Performance is the central theme of the Lava database system.
High row insertion rate
A moderate configuration of server and client will allow insertion of 20,000 rows per second or more. Using exclusive mount rates of 60,000 rows per second or higher are achievable. (Exclusive mount may be used to import large datasets, or to populate a new database from an alternate source).
Very fast seek
Due to the implementation of memory indexes and highly efficient table data buffering, seek times of a few microseconds are achievable.
Very high processing rate
Using conventional row manipulation, sequential data processing can be performed at very high rates. Throughput of 30,000 rows per second is quite easily achieved.
When using the full facilities offered by Lava Distributed SQL, however, it is often not necessary to use conventional row processing. Under many circumstances table data may be processed using array access in the native programming language. In these cases, sequential access of up to several hundreds of thousands of rows per second is possible.
High transaction rate
A very efficient communication layer and threading design optimized for multiple processor cores allow both high numbers of client connections as well as very high transaction rates - processing rates of up to 10 times higher than conventional databases are attainable on identical hardware.