2016 USENIX Annual Technical Conference (USENIX ATC '16)

FPGA-enabled datacenters have shown great potential for providing performance and energy efficiency improvement. In this paper we aim to answer one key question: how can we efficiently integrate FPGAs into stateof- the-art big-data computing frameworks like Apache Spark? To provide a generalized methodology and insights for efficient integration, we conduct an indepth analysis of challenges at single-thread, single-node multi-thread, and multi-node levels, and propose solutions including batch processing and the FPGA-as-a- Service framework to address them. With a step-by-step case study for the next-generation DNA sequencing application, we demonstrate how a straightforward integration with 1,000x slowdown can be tuned into an efficient integration with 2.6x overall system speedup and 2.4x energy efficiency improvement.