Title | ||
---|---|---|
Reproducible Dependability Benchmarking Experiments Based on Unambiguous Benchmark Setup Descriptions |
Abstract | ||
---|---|---|
Dependability benchmarking is performance benchmarking extended by dependability aspects. It is an essential feature of benchmarks of any kind, that they make possible the fair evaluation and comparison of a system's attributes. To avoid biased benchmarking results, independent teams should be able to repeat the benchmark of a system and produce results similar to those of the original benchmark. A semantically unambiguous description of the benchmark setup is a major step towards the goal of reproducible benchmarking experiments. This paper introduces a VHDL-based method for semantically unambiguous benchmark setup description of both the static and dynamic aspects of benchmarks in general. The approach proposed uses VHDL only as unambiguous description and modeling language and does not place restrictions on the actual evaluation of the experiment in any way. A dependability benchmarking experiment using this method is presented. |
Year | DOI | Venue |
---|---|---|
2003 | 10.1109/DSN.2003.1209957 | DSN |
Field | DocType | Citations |
Dependability,Instruction set,Computer science,Dependability benchmarking,Modeling language,VHDL,Videoconferencing,Benchmarking,Distributed computing | Conference | 5 |
PageRank | References | Authors |
0.81 | 1 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Kerstin Buchacker | 1 | 44 | 8.05 |
Mario Dal Cin | 2 | 282 | 40.09 |
Hans-jörg Höxer | 3 | 6 | 1.17 |
Roland Karch | 4 | 16 | 2.99 |
Volkmar Sieh | 5 | 127 | 17.00 |
Oliver Tschäche | 6 | 63 | 5.94 |