GPU computing is everywhere. After some attempts in the antivirus area (see here and here), GPUs are now used for accelerating database SQL queries.
Prior work has shown dramatic acceleration for various database operations on GPUs,
but only using primitives that are not part of conventional database languages such as SQL. This paper implements a subset of the SQLite command processor directly on the GPU.
This paper focuses on accelerating SELECT queries and describes the considerations in an efficient GPU implementation of the SQLite command processor. Results on an NVIDIA Tesla C1060 achieve speedups of 20-70X depending on the size of the result set.
though the finite memory size of the GPU is a significant limitation, allocating just half
of the 4 gigabytes of a Tesla C1060 to store a data set gives the user room for over 134 million rows of 4 integers.
You can download the paper HERE.