Gasgoo Munich- Over the past two years, Tesla, Geely, NIO, and XPENG have successively moved into chip design, turning "in-house silicon" from a niche experiment into an industry obsession. Yet skepticism persists: developing automotive-grade chips is capital-intensive, slow, and technically demanding. Are non-chip automakers crossing over out of genuine necessity, or just to chase a trend?
On May 12, Li Auto Chairman and CEO Li Xiang addressed external criticism that the company's chip ambitions were performative—a costly exercise in trend-following. He avoided empty tech rhetoric and didn't shy away from the controversy. Instead, starting from the practical challenges of deploying AI in the physical world, he clearly broke down the fundamental logic behind Li Auto's push for self-developed silicon.
In-house Chip Is About Solving Pain Points, Not Showing Off
The skepticism surrounding automakers' chip efforts centers on motive and value. Many voices argue that automakers are building chips simply to slap a high-tech label on their brands—appearing cutting-edge while burning resources on projects detached from reality.
Li Xiang's response was direct and unequivocal: "We are developing our own chips not to prove our technical prowess, but to ensure AI can actually function in the physical world. We use in-house silicon to solve problems that current supplier technology cannot overcome." This statement cuts to the core: it is not about flexing technical muscle, but about solving actual problems.
Today, advanced autonomous driving, in-vehicle large models, and cockpit-driving integration rely heavily on chip compute power. While most automakers rely on off-the-shelf generic chips, the shortcomings of these components are becoming glaringly apparent as AI becomes deeply intertwined with mobility scenarios: low compute utilization, high latency, and poor hardware-software coordination make it difficult to meet the rigorous demands of running large models locally and making real-time driving decisions.
Image source: Li Auto
Crucially, these pain points cannot be fully resolved by optimizing supplier solutions, because the architecture of generic chips itself presents fundamental bottlenecks.
Li Xiang added that Li Auto's strategy extends far beyond a single chip; it involves a long-term commitment to more core underlying technologies. In his view, a breakthrough in just one chip cannot support a comprehensive upgrade of the vehicle's AI experience. Only by diving deep into the underlying technology ecosystem can the company break free from external supply chain constraints and avoid being held hostage by critical hardware and algorithms.
This means Li Auto's self-development is not a tentative experiment at a single point, but a systematic strategy where every investment targets genuine technical pain points and user needs—not blind trend-following.
In the AI Era, the Core of Competition Shifts to Holistic Systemic Capability
Having clarified the original intent behind in-house chips, Li Xiang further elaborated on Li Auto's technical logic. He used Apple as a benchmark to deconstruct the rules of competition in the AI era.
"Why does Apple offer the best experience? It's not because one specific technology is the strongest, but because they achieve full-chain autonomous design and responsibility across chips, operating systems, hardware, and cloud services—there can be no weak link." Li Xiang believes that superior user experience stems from closed-loop management and deep hardware-software integration across the entire chain.
He made it clear that Li Auto is learning from Apple not in terms of product form or UI design, but the underlying logic of "soft-hard joint design." "We want to bring AI into the physical world and give everyone an experience like Apple's." To that end, Li Auto is simultaneously advancing full-stack R&D in chips, operating systems, large models, and hardware, building a holistic joint design capability for the artificial intelligence age.
In Li Xiang's view, the logic of industry competition has fundamentally shifted: "The era of competing for single-event championships is over. The AI era is about systemic capability. Only through the joint design of chip architecture, operating systems, models, compilers, hardware design, and production technology—achieving 'all-around excellence'—can one become the champion of user experience."
This assessment aligns with current trends in intelligent vehicles. As AI fuses deeply with the physical world, the competitiveness of smart cars no longer depends on the strength of a single technology, but on the collaborative capability of the entire chain: chips provide compute for algorithms, operating systems act as carriers for the ecosystem, large models optimize interaction and decision-making, and hardware brings all technologies to life.
A weak link in any part of this chain directly impacts user experience. Li Auto's full-stack self-development is a strategic choice aligning with this trend, aiming to eliminate adaptation losses between software and hardware through holistic collaboration, thereby transforming AI technology into tangible mobility value.
Li Xiang's public comments not only provide a clear annotation of Li Auto's in-house chip strategy but also highlight the underlying logic driving the entire smart vehicle industry. Automakers developing their own chips is not the blind cash-burning or hype that outsiders perceive; it is an inevitable choice for addressing technical bottlenecks, seizing industrial initiative, and building differentiated competitiveness.
In the short term, developing underlying technologies requires sustained capital and R&D resources with a long payoff cycle, inevitably inviting skepticism. But in the long run, only by rooting in core underlying technologies and building holistic systemic capabilities can automakers truly shed dependence on external supply chains and achieve technological autonomy.
Looking ahead, as AI and the automotive industry integrate more deeply, software-hardware integration and full-chain self-development will become standard equipment for top-tier automakers.
Under these new competitive rules, only by abandoning the old mindset of isolated breakthroughs, deeply cultivating underlying technologies, and insisting on holistic collaboration can automakers gain a foothold in the second half of the smart mobility race. It is about using solid technical accumulation to create a stable, fluid, and reliable mobility experience for users, propelling China's smart vehicle industry from "application innovation" to "foundational breakthroughs.
