Lessons Learned On Which Applications Benefit When Implemented On Cpu-Fpga Heterogeneous System
2018 INTERNATIONAL CONFERENCE ON EMBEDDED COMPUTER SYSTEMS: ARCHITECTURES, MODELING, AND SIMULATION (SAMOS XVIII)(2018)
摘要
In this work, we provide "lessons learned" from implementing two applications, collision detection and Boolean Gene Regulatory Networks (GRNs) simulation, on a CPU-FPGA heterogeneous platform. Both of these applications have, previously, been implemented and accelerated on FPGA-only devices, but when implemented on a more complete host and co-processor system the additional system factors, such as input and output data communication, impact the results. Using our two applications, we illustrate a set of lessons that need to be considered when porting applications to these emerging heterogeneous systems.
更多查看译文
关键词
Hardware, Collision Detection, Accelarator, Xeon plus FPGA, Verilog
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要