This benchmark is used to test the MPI (and hence interconnect) performance and memory performance of the systems. Although HPCC includes High Performance Linpack (HPL) this benchmark is not currently used on the systems.
HPCC Challenge was downloaded from;
The benchmark was run in two configurations:
- b_eff - MPI (interconnect) performance measures
- STREAM - Memory performance measures
## Analysis
b_eff results are analysed in a Jupyter notebook:
Parsing of the HPCC output file is performed by functions defined in a custom Python module:
STREAM results are read directly from the HPCC output files (links below).