In computer programming, a "PDQ" refers to a parallel data query, which is a process wherein software breaks a query into pieces so several parts of a database can be searched simultaneously. This type of query is optimized for massively parallel processors that work together to increase computer processing speeds needed to run applications that require huge amounts of data.
Instead of one program working on an entire query, multiple programs work on parts of the query simultaneously. For instance, four programs may search four different quarters of data gathered in an entire year, rather than one search for one whole year's worth of quarterly data.
Parallel data queries work best for the creation of large indexes, large tables of organized data, and bulk files that are updated, downloaded or deleted in batches. These types of programs work with hardware that contains multiple processors, has large input/output bandwidth, utilizes CPUs that are not in constant use and accesses sufficient memory. Executing a parallel data query reduces response time for large data warehouses.
Parallelism works on multitudes of levels, including applications that run between operations, computers, servers and even different locations. Parallel data queries are often embedded in computer programming to execute regular searches across huge amounts of data.