Cross-Examining the Computational Performance of Radiation-Tolerant NVIDIA and AMD SoCs

As space missions encounter increasing computational demands, the need for hardware that combines technical capability with resistance to harsh space radiation environments grows. Radiation-hardened (rad-hard) hardware often performs poorly compared to commercial-off-the-shelf (COTS) processors. Consequently, COTS-based systems-on-chips (SoCs) demonstrating some radiation tolerance, such as NVIDIA's Tegra K1 and AMD's Ryzen V1000 series, may provide viable alternatives for low Earth orbit (LEO) missions. This study extends previous individual evaluations by offering a comparative analysis, conducting standardized tests on critical computational components - CPU, GPU, and Single-Instruction Multiple-Data units - across both systems. The comprehensive analysis presented in this study delivers robust comparative performance metrics, equipping the aerospace community with essential data to optimize hardware selection for future space missions.