The US will have less leverage over export control as China becomes more self-sufficient
In the long contest between technological ambition and imposed constraint, Huawei has introduced a theoretical framework it calls the Tau Scaling Law — a proposed path to advanced chip density that would not require the Western tools China has been denied. Announced in late May 2026, the claim suggests that necessity may once again prove the mother of invention, as Chinese engineers seek to route around the bottlenecks that export controls were designed to make permanent. Whether the theory survives contact with the factory floor remains the defining question, but the announcement alone signals that the architecture of technological dependence is being quietly renegotiated.
- Huawei claims its Tau Scaling Law could match 1.4-nanometre chip density by 2031 — without the EUV lithography tools the US has spent years restricting.
- The announcement strikes at the core logic of American export controls, which assumed that denying key equipment would keep China's chip ambitions perpetually out of reach.
- Analysts like Natixis economist Gary Ng warn that if the law proves out, US leverage over China's semiconductor sector could erode significantly and rapidly.
- The moment echoes DeepSeek's 2024 surprise, when a Chinese AI model outperformed expectations built on the assumption that hardware constraints were decisive.
- Skepticism remains grounded: scaling laws have a history of promising more than factories can deliver, and 2031 is still five years and many engineering obstacles away.
Huawei announced a development on Monday that could reframe the global conversation about China's semiconductor capabilities. The company unveiled what it calls the Tau Scaling Law — a theoretical framework it says will allow transistor densities equivalent to 1.4 nanometres by 2031, narrowing the gap with the world's most advanced chipmakers without relying on the tools China cannot access.
The significance lies in what the law promises to circumvent. China's chip ambitions have long been constrained by its inability to obtain extreme ultraviolet lithography machines — the specialized equipment used to etch circuits at the smallest scales. The US has leaned heavily on this dependency, using export restrictions as a lever to slow China's technological rise. Huawei's framework, if it works, would chart a different course — one that does not require Washington's permission.
He Tingbo, who leads Huawei's semiconductor division, framed the announcement as precisely that: a path forward on China's own terms. Analyst Gary Ng of Natixis acknowledged the stakes plainly, noting that a self-sufficient China would leave the US with far less leverage over export controls — while also cautioning that the theory still needs to prove itself in real manufacturing conditions.
The announcement carries echoes of DeepSeek's late-2024 moment, when a Chinese AI model performed at levels many assumed impossible without top-tier American chips — forcing a global recalibration of what constraint actually constrains. Huawei's claim carries similar disruptive weight. But the semiconductor industry has seen promising scaling laws falter between the whiteboard and the wafer. With 2031 still five years away, the question is not only whether the theory holds, but whether it can be built.
Huawei announced a breakthrough on Monday that could reshape how the world thinks about China's ability to make advanced chips without relying on American technology. The company introduced what it calls the Tau Scaling Law—a theoretical framework it says will let it manufacture transistors at a density equivalent to 1.4 nanometres by 2031, closing a significant gap with the world's most advanced chipmakers.
The announcement matters because of what it promises to sidestep. For years, China has struggled to produce cutting-edge semiconductors partly because it cannot easily access extreme ultraviolet lithography tools—the specialized machines needed to etch circuits at the smallest scales. The US has weaponized this dependency, restricting exports of these tools and the advanced chips themselves as a way to constrain China's technological progress. If Huawei's Tau Scaling Law works as described, the company would no longer need those restricted tools to reach the performance levels that define the current frontier of chip design.
He Tingbo, who leads Huawei's semiconductor division and chairs its Scientist Committee, framed the development as a path forward that does not depend on equipment the company cannot obtain. The implication was clear: China could achieve technological parity through a different route, one that does not require permission from Washington.
Analysts are watching closely, though with measured skepticism. Gary Ng, a senior economist at Natixis Corporate and Investment Bank, acknowledged that if the law holds up in real-world testing, it would fundamentally alter the balance of power in the US-China technology competition. "The US will have less leverage over export control as China becomes more self-sufficient," he said. But he also sounded a note of caution: the theory still needs to prove itself in practice. A scaling law that works on paper is not the same as chips that work in factories.
The timing of the announcement echoes an earlier moment that caught global attention. In late 2024, DeepSeek, another Chinese AI company, released a language model that performed at levels many thought impossible without access to the most advanced American chips. That surprise forced a recalibration of assumptions about what China could achieve under constraint. Huawei's Tau Scaling Law announcement carries similar weight—it suggests that Chinese engineers are finding ways around the bottlenecks that were supposed to be insurmountable.
What remains to be seen is whether the theory translates into silicon. The semiconductor industry has a long history of scaling laws that looked promising in research but stumbled when engineers tried to manufacture at scale. The path from 1.4-nanometre density in 2031 is still five years away, and much can change in that time. But if Huawei can deliver on even a fraction of what it is promising, the geopolitical calculus around semiconductors—and with it, the balance of technological power between the US and China—will shift in ways that are only beginning to become clear.
Citas Notables
The US will have less leverage over export control as China becomes more self-sufficient— Gary Ng, senior economist at Natixis Corporate and Investment Bank
La Conversación del Hearth Otra perspectiva de la historia
What exactly is the Tau Scaling Law, and why should anyone outside the chip industry care?
It's essentially a roadmap for how to pack transistors more densely into a chip without using the most advanced manufacturing equipment. The reason it matters globally is that it could let China build world-class semiconductors without depending on American tools or expertise.
So Huawei is saying it can catch up to Intel or TSMC without the machines those companies rely on?
That's the claim. The real question is whether the theory works when you actually try to manufacture millions of chips. Scaling laws look good on paper all the time.
Why is the US so concerned about China having advanced chips in the first place?
Because chips power everything—AI systems, weapons, data centers, communications. If China can make its own advanced chips, it becomes much harder for the US to control what China can and cannot do technologically.
This sounds like it could be another DeepSeek moment, where China surprises everyone with what it can do under pressure.
That's exactly what analysts are thinking. DeepSeek showed that Chinese engineers could work around constraints in ways people didn't expect. If Huawei pulls this off, it suggests that pattern is real.
But you said the law still needs to be tested. What does that actually mean?
It means building actual factories, running actual production lines, and seeing if you can consistently make chips that perform as promised. That's where most ambitious plans fall apart—the gap between theory and manufacturing reality.
When would we know if this works or not?
Huawei is targeting 2031. So we're looking at five years of development before we can really judge whether this changes the game or remains an interesting research project.