Huawei unveils LogicFolding chips to challenge Nvidia in China

Huawei announced plans to roll out advanced smartphone chips this fall using a new engineering method called LogicFolding, a development that comes as American export controls on semiconductors continue to show limited effect in containing Chinese progress. The Chinese company highlighted the approach at an industry event in Shanghai, where it projected that by 2031 its technology could reach performance levels comparable to 1.4-nanometer processes, outpacing current global leaders in some metrics.

This move directly challenges U.S. efforts to restrict access to cutting-edge chipmaking tools, restrictions that were meant to preserve American technological edges but have instead forced Chinese firms to accelerate domestic alternatives. Huawei's earlier Mate 60 phone already demonstrated 5G capabilities through its own advanced silicon, helping the company claw back smartphone market share in China from Apple. Industry observers note that Beijing's support for homegrown semiconductors has intensified exactly because Washington has blocked sales of the most sophisticated equipment.

Nvidia chief Jensen Huang recently acknowledged that his company has effectively yielded the Chinese market for high-end chips to Huawei. American restrictions bar Nvidia from selling its top-tier products there, creating an opening that local players have filled. Analysts tracking the sector point out that this shift raises questions about the long-term impact of export policies crafted in Washington, where concerns over Huawei as a security risk remain high yet results on the ground tell a different story. Supply commitments and revenue forecasts for Nvidia still look strong on paper, but the loss of ground in the world's second-largest economy suggests those gains may rest on shaky foundations.

Broader market data reinforces the narrow base of recent gains. Over the past two years, the S&P 500 posted solid returns, yet stripping out companies tied to AI infrastructure such as Nvidia, AMD, and related suppliers drops performance sharply. Nvidia alone accounts for a sizable slice of the index, with valuations reflecting heavy bets on sustained AI demand. Bank of America analyst Vivek Arya has set a $350 target on Nvidia shares, citing explosive growth from so-called agentic AI applications that require far more compute than basic chat tools. Those projections assume the cycle will remain unprecedented and structural rather than cyclical, but concentration in a handful of names leaves retirement accounts and index funds exposed if competition from China intensifies or if hype outruns actual deployment.

Security questions around advanced AI add another layer. Anthropic has kept its Mythos model, which excels at spotting code vulnerabilities, under tight controls through a limited access program because wider release could arm malicious actors. Governments in Asia have already responded with urgent reviews and patching drives, underscoring how quickly such tools can shift the balance between defense and offense. The same underlying semiconductor capacity that fuels commercial AI also powers these capabilities, meaning U.S. policy choices on chip exports influence not just market share but also who controls the next generation of both commercial and defensive technologies.

American consumers and workers have watched similar patterns before, where restrictions aimed at rivals end up subsidizing foreign innovation while domestic industry absorbs the costs. Huawei's timetable for new chips this fall serves as the latest reminder that technological leadership cannot be maintained through sanctions alone when the underlying industrial base has already been hollowed out by decades of offshoring and elite consensus around global supply chains.