Skip to content

Latest commit

 

History

History
15 lines (8 loc) · 1.52 KB

10. DATABASE ACCELERATION USING RECONFIGURABLE COMPUTING TO ACTUATE THE CPU WORKLOAD.md

File metadata and controls

15 lines (8 loc) · 1.52 KB

DATABASE ACCELERATION USING RECONFIGURABLE COMPUTING TO ACTUATE THE CPU WORKLOAD

The data mining and machine learning have become quite popular these days which has resulted in huge amounts of data. The huge data is due to the prevalent development of mobiles, social networking sites, and IoT products. The efficient management of this data has become a crucial part in deciding the latency involved in processing and computation. The Field Programmable Gate Arrays (FPGAs) are playing a crucial role to accelerate various highperformance applications at data centers. The proposed research work is to develop a reconfigurable device for accelerating database systems by implementing an FPGA as a coprocessor to the existing CPU. The intended work is to utilize the existing memory (Block RAM) and logic blocks on the FPGA. The suggested system yields high throughput by offloading complex database operators from the host CPU workload. As various database queries demand heavy computational runtime the system focus is to lower the query receptivity and optimize the real-time performance. The research aims to execute a reconfigurable accelerator on a Xilinx Zynq 7000 development board. The goal of the project is to have an accelerated throughput of 3.5x compared to existing database accelerators.

Poster Video


Faculty Name

Harsha Vardhan Reddy Vudumula