Topics


The tutorial is organized along the following topics:

  1. DCBench: A Bechmark Suite for Data Center Workloads.
    • a) Background and motivation
    • b) What are the differences between DCBench and CloudSuite?
    • c) Overview of the workloads in our benchmarks.
    • d) Workloads’ application scenarios
    • e) How to configure and run the benchmarks on the real systems.
    • f) Use cases: workload analysis in Data centers.
  2. Big Data Benchmarks
    • a) The background and motivation
    • b) Our methodology of benchmarking big data systems.
    • c) The Chosen four benchmark in Search Engine Systems.
    • d) Workloads’ application scenarios
    • e) How to configure and run the big data benchmarks on the real systems.
    • f) Use cases: Introducing the runtime properties and workload characterization.
  3. Cloud benchmark suites
    • a) Characterizing and benchmarking cloud systems running data processing applications (CloudRank-D).
    • b) Instroduction of CloudRank benchmark suites
    • c) How to configure and run CloudRank on the real systems.
    • d) Use case: Performance Evaluation Of a Cloud System running data processing applications.
    • e) Characterizing and benchmarking cloud systems running desktop environments (CloudRank-V)
    • f) How to configure and run CloudRank-V
    • g) Use case: Performance Evaluation Of Desktop Clouds
  4. A Benchmark Factory for GPGPUs
    • a) the difficulties of developing GPGPU programming
    • b) the need for GPGPU benchmarks
    • c) Why a factory for GPGPU benchmarks?
    • d) Overview of the GPGPU benchmark factory
    • e) Evaluation of the automatically generated benchmarks
    • f) On going research