The collapse phenomenon, the parallelism principle and states correlation are used to define a type of a Grover rapid search engine. In our approach, the observer’s query and the Grover-unsorted-data are stored in different memories where the global state is represented by a tensor product of the associated states. In the proposed formalism, each query-state input activates an adjusted operator that implements the unsorted state in an appropriate 2-D Grover representation. It will be shown that once the representation is set, it takes mainly two operations to complete the whole query search. This seems to be a very efficient search algorithm.