Cloud Computing. Third International Conference, CloudComp 2012, Vienna, Austria, September 24-26, 2012, Revised Selected Papers

Research Article

Performance Evaluation of Embedded Processor in MapReduce Cloud Computing Applications

Download
439 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-03874-2_5,
        author={Christoforos Kachris and Georgios Sirakoulis and Dimitrios Soudris},
        title={Performance Evaluation of Embedded Processor in MapReduce Cloud Computing Applications},
        proceedings={Cloud Computing. Third International Conference, CloudComp 2012, Vienna, Austria, September 24-26, 2012, Revised Selected Papers},
        proceedings_a={CLOUDCOMP},
        year={2014},
        month={6},
        keywords={cloud computing green computing embedded processors mapreduce data centers},
        doi={10.1007/978-3-319-03874-2_5}
    }
    
  • Christoforos Kachris
    Georgios Sirakoulis
    Dimitrios Soudris
    Year: 2014
    Performance Evaluation of Embedded Processor in MapReduce Cloud Computing Applications
    CLOUDCOMP
    Springer
    DOI: 10.1007/978-3-319-03874-2_5
Christoforos Kachris1,*, Georgios Sirakoulis1,*, Dimitrios Soudris2,*
  • 1: Democritus University of Thrace
  • 2: National Technical University of Athens
*Contact email: ckachris@ee.duth.gr, gsirak@ee.duth.gr, dsoudris@microlab.ntua.gr

Abstract

Current data centers consume huge amount of power to face the increasing network traffic. Therefore energy efficient processors are required that can process the cloud applications efficiently without consuming excessive power. This paper presents a performance evaluation of the processors that are mainly used in high performance embedded systems in the domain of cloud computing. Several representative applications based on the widely used MapReduce framework are mapped in the embedded processor and are evaluated in terms of performance and energy efficiency. The results shows that high performance embedded processors can achieve up to 7.8x better energy efficiency than the current general purpose processors in typical MapReduce applications.