How Long Will CAN Stick Around As Rival Networks Speed Up?

As automotive systems evolve toward software-defined architectures, in-vehicle networks are undergoing a significant transformation. While legacy protocols such as CAN and LIN remain entrenched due to... Read original →

Automotive Ethernet’s ascent is triggering a paradigm shift in in-vehicle communication stacks. Technically, TSN and IEEE 802.1AS are forcing deep integration of SerDes, MACsec, and PTP clocking into automotive SoCs, compelling Rambus and Infineon to redesign PHY-layer IP—while legacy CAN/LIN controller markets face >30% contraction within two years, squeezing smaller MCU vendors. On compliance, EU GSR 2 and UN-R155 mandate intrusion detection capabilities, exposing CAN’s lack of native encryption and pushing OEMs to overhaul ECU security architectures, inflating BOM costs by 5–8%. Strategically, Keysight and Siemens EDA are bundling validation toolchains around software-defined vehicles, while Imagination bets on MIPI A-PHY-to-Ethernet bridging to bypass legacy AUTOSAR constraints. Over the next 12–24 months, as NVIDIA Thor and Qualcomm Snapdragon Ride platforms deploy zonal architectures, Ethernet will cascade from domain controllers to zone nodes, eroding CAN’s foothold even in non-critical functions—with 2027 marking the tipping point for protocol obsolescence.