Abstract
The HEPiX Benchmarking Working Group has developed a framework to benchmark the performance of a computational server using the software applications of the High Energy Physics (HEP) community. This framework consists of two main components, named HEP-Workloads and HEPscore. HEP-Workloads is a collection of standalone production applications provided by a number of HEP experiments. HEPscore is designed to run HEP-Workloads and provide an overall measurement that is representative of the computing power of a system. HEPscore is able to measure the performance of systems with different processor architectures and accelerators. The framework is completed by the HEP Benchmark Suite that simplifies the process of executing HEPscore and other benchmarks such as HEP-SPEC06, SPEC CPU 2017, and DB12. This paper describes the motivation, the design choices, and the results achieved by the HEPiX Benchmarking Working group. A perspective on future plans is also presented.
Highlights
The High Energy Physics (HEP)-SPEC06 (HS06) benchmark [1], based on SPEC CPU 2006 [2], is currently used by the Worldwide LHC Computing Grid (WLCG) [3] community to estimate the performance of a computing server
This section describes the use of the HEP Benchmark Suite with the HS0664bits and HEPscore benchmarks
The benchmarks were run on servers at a number of WLCG and HPC sites used by the HEP community to evaluate a wide range of CPU models
Summary
The HEP-SPEC06 (HS06) benchmark [1], based on SPEC CPU 2006 [2], is currently used by the Worldwide LHC Computing Grid (WLCG) [3] community to estimate the performance of a computing server. HS06 has been used for over a decade, satisfying the WLCG requirements in a landscape that progressively evolved from CPUs with a few cores to multi-cores CPUs. When HS06 was established, the HEP applications shared several commonalities with the SPEC CPU 2006 workloads included in HS06: they were characterized by singlethreaded and single-process applications, compiled in 32-bit mode, with a memory footprint of about 1 GB per process. Even if HS06 may continue to be a viable benchmark for evaluating the performance of x86 CPUs, the community will soon require a benchmark to evaluate
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
