Technology & Science
Huawei Unveils τ-Scaling Law and LogicFolding to Target 1.4 nm-Class Density by 2031
On 25 May 2026, Huawei’s He Tingbo introduced the τ (Tau) Scaling Law and companion LogicFolding chip architecture at IEEE ISCAS in Shanghai, pledging commercial Kirin chips this fall and 1.4 nm-equivalent transistor density by 2031—an attempt to sidestep Moore’s geometric-shrinking limits using time-constant optimisation.
Focusing Facts
- Huawei says it has already designed and mass-produced 381 chips under the τ Scaling Law over the past six years.
- The first LogicFolding-based Kirin SoC, due in late 2026, is touted to raise transistor density 53.5 % to 238 MTr/mm² and cut P-core power by 40 %.
- Company roadmap claims τ-based high-end chips will reach a 1.4 nm-equivalent density by 2031 despite current export controls on EUV tools.
Context
Every few decades the semiconductor industry hits a wall—1965’s end of vacuum-tube “scaling”, the 2004 power-wall that birthed multicore, or the 2011 shift to FinFETs—all forcing designers to seek new dimensions beyond raw lithographic shrink. Huawei’s τ-Scaling echoes IBM’s 1997 adoption of copper interconnects: an architectural workaround when materials physics stalled feature-size progress. Long-term, it signals two entwined trends: (1) the global race to extend performance without EUV, especially under U.S. sanctions that bar Huawei from ASML’s high-NA tools; and (2) the gradual decoupling of design innovation from Western fabs, reminiscent of Japan’s 1980s DRAM ascendance before U.S. trade pressure reshaped the market. If Huawei can translate “equivalent” density claims into mass-produced silicon—still a huge ‘if’ given foundry partner SMIC’s current 5–7 nm ceiling—it could redistribute technological leadership and supply-chain leverage for decades. If it fails, τ-Scaling may be filed alongside other optimistic post-Moore proposals (quantum dot cellular automata, 1997; 3-D IC hype, 2008). Either way, the announcement underscores that the century-long narrative of computing progress is migrating from pure lithography to multidimensional system-level optimisation.
Perspectives
Chinese and China-aligned business media
e.g., South China Morning Post, FoneArena — Present Huawei’s Tau Scaling Law as a genuine breakthrough that will let the firm keep boosting chip performance despite Moore’s Law limits, underscoring the number of chips already mass-produced and the lofty 1.4 nm goal. Coverage echoes Huawei’s promotional language and national techno-optimism, offering little independent verification of the engineering claims while framing them as proof of China’s high-tech resilience under U.S. sanctions.
Financial and investor-focused outlets
e.g., Investing.com — Highlight the commercial roadmap—Kirin, Ascend and Kunpeng chips—and frame the Tau Scaling Law chiefly as a signal that Huawei could rival Nvidia and AMD by 2031, useful information for market watchers. The angle prioritises future revenue potential and stock-market relevance, so it tends to relay Huawei’s projections uncritically and downplays technical skepticism in favour of forward-looking numbers.
Western tech trade press with a skeptical stance
e.g., The Register — Covers Huawei’s announcement with noticeable doubt—calling the new rule a “so-called law,” stressing “if accurate,” and noting it merely ‘proposes’ replacing Moore’s Law rather than proving it. The Register’s trademark snark and geopolitical caution may lead it to dismiss or underplay Huawei’s advances, potentially reflecting broader Western skepticism toward Chinese tech claims.
Like what you're reading?