Dataset of execution times of programs compiled under different GCC optimizaion levels (-O0, -O1, -O2, -O3) on the Xilinx Zynq UltraScale+ MPSoC ZCU104 (Cortex A53)
收藏NIAID Data Ecosystem2026-05-02 收录
下载链接:
https://zenodo.org/record/13885210
下载链接
链接失效反馈官方服务:
资源简介:
This dataset contains execution time measurements for a set of programs from the TACLe benchmark suite, including lift, powerwindow, bsort, matrix1, epic, dijkstra, mpeg2 and quicksort.
The data was collected using RT-Bench, an open-source framework designed to execute these benchmarks periodically in a real-time context.
Each program was compiled using five different GCC optimization levels (-O0, -O1, -O2, -O3, and -Ofast) and executed periodically 1000 times.
The experiments were conducted on the Zynq ZCU104 platform, which features a quad-core Arm Cortex A53 processor, running the Ubuntu 22.04 LTS arm64 OS with the 5.15.0-1015-xilinx-zynqmp kernel.
This dataset is published at ETFA 2024 as part of a work-in-progress paper called "Impact of compilation optimization levels on execution time variability"
创建时间:
2024-10-08



