Simon developed the Commbench experiment framework for the LLNL/benchpark repository, enabling standardized benchmarking of communication libraries in high performance computing environments. Leveraging Python, CUDA, and ROCm, Simon designed Commbench to support reproducible experiments across diverse hardware platforms. He also integrated the El Dorado system configuration, broadening Benchpark’s compatibility with advanced computing resources and streamlining experimental setup. This work addressed the need for consistent performance evaluation in HPC, allowing researchers to compare results across different architectures. Simon’s contributions demonstrated depth in benchmarking and system integration, enhancing Benchpark’s utility for the HPC community and supporting more rigorous, hardware-agnostic performance analysis.