Huawei Says It Found a Way Around US Chip Bans — Nvidia and Intel Should Take Notice

TLDR

• Huawei says it can match 1.4-nanometer chip performance by 2031 without ASML machines
• The company calls its approach the “Tau Scaling Law,” focusing on signal speed and circuit stacking
• New “LogicFolding” architecture will debut in Kirin smartphone chips this fall
• Huawei has already mass-produced 381 chip models using this method over six years
• Nvidia CEO Jensen Huang says Huawei is a strong competitor and Nvidia has “largely conceded” China’s AI chip market

---

Huawei says it has found a way to build chips that match the world’s best — without the equipment it’s been blocked from buying.

The Chinese tech company announced at a conference in Shanghai on Monday that it expects to design chips by 2031 with transistor density equal to a 1.4-nanometer process. That’s the next frontier in chip performance, currently being targeted by Intel, TSMC, and Samsung.

The catch? Those companies use machines made by Dutch firm ASML to reach that level. Huawei has been blocked from accessing that equipment due to U.S. export restrictions.

Huawei’s chip arm president, He Tingbo, presented what the company calls the “Tau Scaling Law” at the 2026 IEEE International Symposium on Circuits and Systems. The idea is to improve chip performance by reducing how long it takes data to move through a chip, rather than just making transistors smaller.

“Our solution is feasible and affordable,” He said at the event.

What Is LogicFolding?

Huawei’s new engineering method, called “LogicFolding,” stacks multiple layers of circuits within a single chip. The company says this shortens internal wiring and improves performance. The upcoming Kirin smartphone chips, set to launch this fall, will be the first to use this architecture.

Huawei says it has already mass-produced 381 chip models using related techniques over the past six years, for use in smartphones and artificial intelligence systems.

The company did not provide independent performance data to back up its claims.

What This Means for Nvidia

Huawei’s progress puts more pressure on Nvidia in China. Nvidia CEO Jensen Huang said earlier this month that Nvidia’s market share in China dropped from around 95% to nearly zero following years of U.S. export restrictions.

Huang told CNBC the company has “largely conceded” China’s AI chip market to Huawei, calling it a “very strong competitor.”

Nvidia has received U.S. licenses to sell its H200 AI chips in China, but Chinese regulators have not yet approved those sales. Huang said he still wants to compete there. “It would be terrific to serve that market,” he said.

George Chen of The Asia Group said the window for Nvidia to sell advanced AI chips in China is narrowing as Huawei grows stronger locally.

Nvidia stock currently holds a Strong Buy consensus on Wall Street, based on 39 Buy ratings, one Hold, and one Sell over the past three months. The average price target is $302.61, implying roughly 40% upside from current levels.

Challenges Still Ahead

Analysts say Huawei’s approach is promising but not proven at scale. The circuit-stacking method faces issues like overheating and requires more complex software to coordinate chip layers.

Huawei only achieved stable results with the new technology in the past year, according to people familiar with the progress. The company still needs to work with data centers and equipment vendors to confirm it works in large-scale applications.

Analyst Lian Jye Su of Omdia said: “Whether Huawei will gain a distinct advantage here remains to be seen, but it’s at least an alternative path forward.”

If Huawei delivers on its 2031 target, it could change assumptions about what’s possible in chip manufacturing without cutting-edge Western equipment.