High-speed Ethernet has evolved into a foundational communication backbone in modern distributed systems, from traditional high-performance computing environments to emerging automotive platforms. Research over the past two decades has examined Ethernet’s latency characteristics, bandwidth scalability, electromagnetic compatibility constraints, energy considerations, and security limitations under increasing system complexity. Early comparative studies such as the evaluation of Ethernet versus Myrinet for message-passing interface (MPI) communication demonstrated the tension between low-latency interconnects and commodity-grade networking technologies, motivating Ethernet’s evolution from 1 Gb/s to 10 Gb/s architectures (Mujumder & Rixner, 2004). Subsequent improvements in 10-Gigabit Ethernet—including TCP offload engines, optimized NIC designs, and commodity hardware acceleration—have significantly improved end-to-end throughput (Feng et al., 2005; Maeda et al., 2018). Yet electrical and optical domains present challenges, particularly under real-world electromagnetic interference (EMI), bandwidth limitations of copper-based physical layers, and the performance trade-offs between modulation formats (Kolahi & Soorty, 2011; Browning et al., 2011).

Parallel to these performance advancements, the rise of automotive Ethernet for advanced driver-assistance systems (ADAS), camera networks, and in-vehicle communication created new security and functional-safety concerns. The survey of automotive Ethernet vulnerabilities highlights risks such as spoofing, replay attacks, and protocol-layer compromise (Douss et al., 2023). Emerging countermeasures include TLS-based in-vehicle authentication (Zelle et al., 2017), privacy-preserving data exchange (Mi et al., 2018), and enhanced controller-area network (CAN-FD) security architectures (Woo et al., 2016; Lin & Sangiovanni-Vincentelli, 2012). Furthermore, automotive EMC standards such as CISPR-25:2021 outline measurement limits to safeguard sensitive on-board receivers, prompting system designers to adopt shielding solutions validated through simulation frameworks such as HyperLynx (Karim, 2025).

This article synthesizes the evolution of high-speed Ethernet through the lenses of performance, EMI resilience, hardware design, and cyber-security, offering a deep theoretical framing of how these factors collectively shape modern distributed and automotive communication networks. The extensive discussion identifies gaps, practical system design considerations, and a forward-looking perspective for high-speed and automotive Ethernet research.