Arteris, announced that its system IP technology has been deployed by Li Auto Inc. in the proprietary autonomous driving systems-on-chip (SoCs) of the L9 Livis high-tech flagship SUV. Arteris FlexNoC 5 network-on-chip (NoC) interconnect IP and Magillem SoC integration automation software are supporting AI compute data movement and integration automation for Li Auto’s current and future smart vehicles.
Li Auto develops extended-range electric vehicles (EREVs), battery electric vehicles (BEVs), and smart vehicle platforms that incorporate advanced driver assistance systems (ADAS), high-performance compute, and in-vehicle experiences. The company developed an in-house autonomous driving SoC that delivers 2,560 trillion operations per second (TOPS) for AI computational tasks.
Li Auto used Arteris FlexNoC 5 and Magillem software to meet high-performance compute needs, energy efficiency targets, and functional safety requirements. The technology enabled efficient AI data movement within the SoC through optimized interconnect architectures that balance performance, power, and area (PPA). It also accelerated time to production for the current implementation and future derivative designs.
“Leveraging the Arteris physical awareness capabilities of FlexNoC 5 silicon IP, Li Auto quickly converged on optimal architecture in the 5 nm process node, while reducing design risk and improving system-level efficiency,” the announcement stated. The solution supported the integration of diverse processing elements, including CPUs, GPUs, AI accelerators, sensor interfaces, and functional safety capabilities.
“We chose Arteris for our SoC development to help us meet the high functional, performance, and safety requirements for autonomous driving chips and to ensure a great user experience in our smart vehicles,” said Luo Min, Head of Computing Unit at Li Auto.
“Automotive innovation is being driven by a fundamental shift toward centralized compute and AI-defined functionality,” said K. Charles Janac, president and CEO of Arteris. “Its choice to work with Arteris to ensure optimized compute for their vehicles reflects the growing importance of silicon data movement as a critical factor meeting the performance, power, safety and security requirements of today’s physical AI systems.”
Arteris products are used in systems ranging from AI data centers to edge AI devices and physical AI systems, including autonomous vehicles.
