Hitachi

Hitachi data-processing technology yields 100-fold improvement in data analytics speeds

Hitachi has developed a database management system that incorporates hardware (FPGA*1) internal memory for high-speed data processing and technology that achieves high-speed parallel processing of data within the FPGA. This approach achieves data analytics rates up to two orders of magnitude (100-fold) faster than conventional methods.

Enables high-speed processing in response to growing demand for self-service analytics

Self-service analytics has attracted growing attention in recent years because it allows users to process analytics for large data sets instead of contracting such work out to data scientists or other specialists. As one example of self-service analytics, a teller at a client bank can enter information provided by a customer to an analytics system on the spot, making it possible to recommend tailored products in real time. A major hurdle to such self-service analytics has been the data processing performance of real-time data analytics systems. Targeting the development of database management systems and parallel processing technologies as solutions to this challenge, Hitachi has boosted the speed of data analytics up to 100-fold. In addition, Hitachi has developed a prototype real-time data analytics system that integrates the two newly developed technologies with Pentaho Business Analytics*2 business analytics software and flash data storage.*3 Hitachi is seeking to further accelerate data analytics processing and to contribute to business efficiency improvements through the validation of real-time data analytics system in real world business applications.

*1
FPGA (Field Programmable Gate Array)
An integrated circuit manufactured to be programmable by the purchaser. In general, FPGA is inexpensive compared to application specific circuits.
*2
Pentaho Business Analytics
Software to specify how to analyze the data and view the results.
*3
Flash storage
Storage equipment employing flash memory as the storage device.

Published: August 3, 2016