Worked on the intel/compute-benchmarks repository to address a critical issue in performance metric reporting for memory benchmarks. Focused on low-level programming in C++, the developer identified and corrected a bug where the random access noop unit was reporting time in nanoseconds instead of the intended memory bandwidth in gigabytes per second. This targeted fix involved applying and validating a patch to ensure accurate metric output, directly improving the reliability of benchmark data used for engineering optimization and capacity planning. The work demonstrated strong debugging skills and attention to detail in performance benchmarking, contributing to more trustworthy results for downstream engineering teams.