Method and apparatus for record pivoting to accelerate processing of data fields
First Claim
Patent Images
1. An apparatus comprising:
- a processor;
a memory comprising a first buffer, a second buffer, and a third buffer; and
a member of the group consisting of (1) a graphics processor unit (GPU), (2) a multi-core processor, (3) a cell processor, and (4) a reconfigurable logic device;
wherein the processor is configured to (1) receive a plurality of incoming records, the incoming records comprising a plurality of fields, (2) write the incoming records to the first buffer, (3) identify which of the fields of the incoming records are to be processed by the member, and (4) write the identified fields to the second buffer;
wherein the member is configured to (1) read the fields in the second buffer, (2) perform a data processing operation on the read fields to generate a plurality of processed fields, and (3) write the processed fields to the third buffer; and
wherein the processor is further configured to (1) read the processed fields in the third buffer, (2) read the fields for the incoming records in the first buffer, and (3) merge the processed fields with the fields from the incoming records to generate a plurality of outgoing records such that the outgoing records include the processed fields in place of their corresponding fields of the incoming records.
1 Assignment
0 Petitions
Accused Products
Abstract
Various methods and apparatuses are described for performing high speed translations of data. In an example embodiment, record layout detection can be performed for data. In another example embodiment, data pivoting prior to field-specific data processing can be performed.
345 Citations
33 Claims
-
1. An apparatus comprising:
-
a processor; a memory comprising a first buffer, a second buffer, and a third buffer; and a member of the group consisting of (1) a graphics processor unit (GPU), (2) a multi-core processor, (3) a cell processor, and (4) a reconfigurable logic device; wherein the processor is configured to (1) receive a plurality of incoming records, the incoming records comprising a plurality of fields, (2) write the incoming records to the first buffer, (3) identify which of the fields of the incoming records are to be processed by the member, and (4) write the identified fields to the second buffer; wherein the member is configured to (1) read the fields in the second buffer, (2) perform a data processing operation on the read fields to generate a plurality of processed fields, and (3) write the processed fields to the third buffer; and wherein the processor is further configured to (1) read the processed fields in the third buffer, (2) read the fields for the incoming records in the first buffer, and (3) merge the processed fields with the fields from the incoming records to generate a plurality of outgoing records such that the outgoing records include the processed fields in place of their corresponding fields of the incoming records. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33)
-
Specification