Evolution of semiconductor-powered ADAS Systems of Tesla and BYD: A guide for ADAS chip design engineers

Advanced Driver Assistance Systems (ADAS) have transformed the automotive industry, with Tesla and BYD emerging as global leaders in electric vehicle (EV) and self-driving innovation. In March 2015, EEHerald celebrated Tesla’s Model S for its over-the-air (OTA) updates enabling autonomous features, a milestone that set the stage for a decade of progress. Today, Tesla’s Full Self-Driving (FSD) and BYD’s God’s Eye (DiPilot) systems showcase divergent approaches to ADAS, powered by advanced semiconductor chips. This article, tailored for ADAS engineers, traces the evolution of Tesla and BYD’s self-driving technologies from 2015 to May 18, 2025, with a focus on the semiconductor architectures driving their systems, comparing their design, performance, and implications for future development.

The 2015 Starting Point: Tesla’s Semiconductor Foundation:

In 2015, Tesla’s Model S introduced Autopilot, a Level 2 ADAS system, using Nvidia’s Tegra 3 processor (quad-core Cortex-A9, 40nm process) alongside radar, 12 ultrasonic sensors, and a forward-looking camera. The Tegra 3, delivering ~1.3 TFLOPS, supported basic lane-keeping and adaptive cruise control, marking Tesla&rsqu...