While the global debate on artificial intelligence is consumed by headlines about bans on Chinese open-source models and the chip war between Washington and Beijing, an industrial giant is quietly executing a strategy that could redefine who controls the next technology cycle. Lenovo, the world’s largest PC maker, has invested in over a hundred AI companies through its venture capital arm, Lenovo Capital. But it’s not chasing the next foundational model startup. Its investment chief, Song Chunyu, has described the approach as that of a “sniper”: surgical investments in robotics, coding agents, and computing infrastructure. This bet reveals a concrete geopolitical reading: value no longer lies in the model, but in the application layer and physical control of the entire stack.
The Sniper Tactic: Why Lenovo Capital Avoids the Model War
Song Chunyu, Senior Partner and CIO of Lenovo Capital, told SCMP Tech that the investment arm is “doubling down” on the next industrial cycle, expanding beyond foundational models. The “sniper” metaphor is no accident: while other venture capital firms cast wide nets over hundreds of generative AI startups, Lenovo Capital selects specific targets in segments where the hardware giant can bring real industrial value.
The decision makes business sense. Foundational models—like those from OpenAI, Anthropic, or Chinese players Qwen and DeepSeek—require multi-billion-dollar investments in computing and talent, with uncertain returns and fierce competition. Lenovo, with its expertise in hardware manufacturing, global supply chains, and enterprise distribution, doesn’t need to own the best model; it needs to control the layers where its competitive advantage is real: chips, servers, data centers, and increasingly, physical robots and software agents that automate business processes.
The portfolio of over one hundred companies spans robotics, coding agents, computing infrastructure, chips and hardware, as well as models and applications. It is a map of the entire value chain, from silicon to software that writes code. For the global audience of investors and policymakers, the signal is clear: a Chinese giant is betting on sovereignty of the complete tech stack, not the hegemony of a single model.
Robotics and Coding Agents: Physical and Logical Control of the Next Cycle
Two sectors stand out in Lenovo Capital’s strategy: robotics and coding agents. These are not random choices. Robotics represents physical control of the world: industrial arms, autonomous vehicles, drones, and warehouse automation systems. China already dominates much of the global industrial robot manufacturing, with companies like Siasun, Estun Automation, and Lenovo’s own robotics division competing in the global market.
Coding agents, meanwhile, represent logical control: AI systems that write, review, and optimize software code. Companies like GitHub Copilot (Microsoft) and Codeium have shown that coding agents can dramatically boost developer productivity. But Lenovo Capital is not investing in the next productivity tool for programmers; it is betting on agents that can integrate into its own manufacturing and distribution processes, cutting costs and speeding up industrial software development.
Song Chunyu has not disclosed the specific names of the startups Lenovo Capital has invested in, but the focus on these two sectors suggests a clear investment thesis: the next industrial cycle will be won not by the company with the best language model, but by the one that can deploy AI in the physical world—robots that assemble servers—and the logical world—agents that write the code controlling those robots.
Computing Infrastructure: The Quiet Bet on Chips and Data Centers
The third pillar of Lenovo Capital’s strategy is computing infrastructure. Here, the Chinese giant is playing on home turf. Lenovo is one of the world’s largest server manufacturers, competing directly with Dell, HPE, and Supermicro. Its infrastructure division, Lenovo ISG (Infrastructure Solutions Group), supplies servers, storage, and high-performance computing solutions to enterprises and governments worldwide.
Investing in computing infrastructure startups makes strategic sense: the more demand grows for AI computing, the more servers, GPUs, and cooling systems the market needs. Lenovo not only manufactures these systems but can also integrate startup technologies into its own products, creating a virtuous cycle of investment and commercialization.
The bet on infrastructure also has geopolitical implications. With U.S. export restrictions on advanced chips to China, companies like Lenovo need to diversify their supply sources and develop domestic capabilities. Investing in Chinese startups focused on chips and computing hardware is a way to secure access to critical technologies without relying solely on U.S. suppliers like NVIDIA or AMD.
What Does This Mean for the Global Market?
Lenovo Capital’s strategy sends several signals to global markets. First: AI is not just a model war; it is a war over infrastructure and applications. While headlines focus on whether DeepSeek surpasses GPT-4 or whether European regulators will ban Chinese models, the savviest companies are placing bets across the entire value chain.
Second: China is not abandoning the AI race; it is pivoting to areas where it has real advantages. Manufacturing, logistics, industrial automation, and hardware-software integration are traditional strengths of the Chinese economy. Lenovo Capital is betting that these strengths, supercharged by AI, will generate the next wave of growth.
Third: Lenovo’s “sniper” strategy could become a model for other industrial giants. Companies like Siemens, Bosch, Foxconn, or even Samsung could follow a similar pattern: investing in startups that complement their existing capabilities, while avoiding direct competition with Silicon Valley giants on foundational models.
Future Outlook: The Next Industrial Cycle Will Be About Integration, Not Models
The most important lesson from Lenovo Capital’s strategy is that value in AI is shifting. Over the past two years, the market has been obsessed with foundational models: who trains the largest, who reaches AGI first, who raises the most funding. But the history of technology shows that real value often lies in the application and distribution layer, not the basic infrastructure layer.
Lenovo, with its global manufacturing expertise, supply chains, and enterprise distribution, is positioning itself to capture that value. It doesn’t need to be the next OpenAI; it needs to be the company that integrates AI into every server, every robot, and every software agent it sells to its enterprise customers.
For investors, policymakers, and professionals tracking the geopolitics of AI, the signal is unmistakable: the next industrial cycle will be decided not by labs in Silicon Valley or Beijing, but by factories, data centers, and production lines where AI turns into real value. And Lenovo, with its sniper strategy, is aiming directly at that future.