High-Level Synthesis Case Study: Implementation of a Memcached Server

Abstract: High-Level Synthesis (HLS) aspires to raise the level of abstraction in hardware design without sacrificing hardware efficiency. It has so far been successfully employed in signal and video processing but has found only limited use in other areas. This paper utilizes a commercial HLS tool, namely Vivado(R) HLS, to implement the processing of a common data center application, the Key-Value Store (KVS) application memcached, as a deeply pipelined dataflow architecture. We compared our results to a fully equivalent RTL implementation done previously in our group and found that it matches its performance, yields tangible improvements in latency (between 7-30%) and resource consumption (22% in LUTs and 35% in registers), all while requiring 3x less lines of code and 2x less development time. The implementation was validated in hardware on a Xilinx(R) VC709 development board, meeting timing requirements for 10Gbps line rate processing.