惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

量子位
Recent Announcements
Recent Announcements
D
Docker
V
V2EX
阮一峰的网络日志
阮一峰的网络日志
Vercel News
Vercel News
Microsoft Security Blog
Microsoft Security Blog
The GitHub Blog
The GitHub Blog
U
Unit 42
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
月光博客
月光博客
腾讯CDC
B
Blog
博客园_首页
罗磊的独立博客
D
DataBreaches.Net
IT之家
IT之家
酷 壳 – CoolShell
酷 壳 – CoolShell
L
LangChain Blog
aimingoo的专栏
aimingoo的专栏
MongoDB | Blog
MongoDB | Blog
GbyAI
GbyAI
Stack Overflow Blog
Stack Overflow Blog
M
MIT News - Artificial intelligence

KrASIA

GoodMe takes ready-to-drink beverages beyond its own stores Mi Liangchuan’s exit deepens pressure on Xpeng’s robotics push Surging users, widening losses, and leased compute: Behind SiliconFlow’s IPO filing BYD spots North America foothold amid US-Canada tariff discord Momenta makes public debut in Hong Kong, puts physical AI in spotlight IPO aspirant SAIC Mobility grows orders, but platforms capture much of the upside Chinese automakers overtake Japanese rivals in Europe despite EV tariffs Li Auto cuts out middle layer in latest R&D restructuring Growatt’s energy storage shift drives third bid for Hong Kong IPO Deals in brief: Vynn Capital to finance Etaily’s expansion, CATL backs CarbonScape as partner, 17 China investments, and more UBTech’s UWorld U1 tests demand for humanoid robots at home 31-year-old founder steers Direct Drive Tech toward Hong Kong IPO Asia's AI rally winners face a rising leverage problem Can a smart ring make health monitoring part of daily life? Chinese-owned Lotus bets on hybrid SUV for US turnaround Chinese door-to-door logistics networks grow in US to counter trade war Hong Kong's hub ambitions are not a zero-sum game Volcano Engine bets better models, not lower prices, will decide the MaaS race Singapore moves ahead with autonomous taxis as trials expand Hong Kong courts GBA-ASEAN connector role as trade and investment ties grow Deals in brief: Airwallex raises Series H funding, Igloo acquires Eazy Digital, nine China investments, and more CaoCao Mobility’s RoboX plan targets autonomous driving’s next phase Chinese appliance brands gain ground in Southeast Asia Robotics will not get its “GPT moment” by following LLMs, Agibot chief scientist says Flanked by DJI and Insta360, this company has grown 50% a year for five straight years Hong Kong pushes Beijing's bay area dream as it advances first five-year plan Asian streaming services go big on microdramas and AI content RayNeo extends AR glasses lead as smart eyewear demand shifts Huawei lays out six priorities for AI-era mobile networks Seer Robotics rides 24-hour market swing in Hong Kong debut
Huawei says new Kirin chip for phones overcomes US clampdown
Nikkei Asia · 2026-06-02 · via KrASIA

Huawei Technologies said on May 25 it has found a new way to design chips to bring its semiconductor capabilities close to those of global chipmakers Taiwan Semiconductor Manufacturing Company (TSMC) and Intel, as China’s tech champion continues working to overcome US sanctions that have cut it off from top-tier global suppliers.

Huawei said that over the past six years, it has developed a new principle, known as the Tau scaling law, to break through the limits of Moore’s Law and keep pursuing advanced semiconductor manufacturing technologies, He Tingbo, Huawei’s semiconductor chief, said at the 2026 IEEE International Symposium on Circuits and Systems, or ISCAS, in Shanghai.

The Tau scaling law is also called “Her’s law” by Huawei, in recognition of He, who joined the company in 1996 and has been leading Huawei’s efforts to develop alternative ways to design chips amid Washington’s persistent crackdown. As president of its semiconductor business department, she has overseen the transformation of chip arm HiSilicon Technologies from a small design team into a globally recognized designer of chips for smartphones, PCs, artificial intelligence computing, and more.

Based on this new principle, He said, engineers can reduce signal delay and steadily improve transistor density to keep advancing chip computing power beyond the limits of Moore’s law, which relies on shrinking transistors ever smaller to increase computing power.

Speaking at the conference, He described how Her’s law led to the development of Huawei’s new LogicFolding chip architecture, which sharply increases chip performance. She cited the Kirin system-on-chip chipset for phones, due out later this year, as an example, saying it delivers a 55% increase in transistor density and a 41% power-efficiency compared with traditional designs.

These gains were “obtained not through a new lithography step but through a topological reorganization of the spatial distribution of logic in three dimensions,” according to He.

Huawei believes its new approach to chip development will become an industry standard. “Six years ago, the geometric roadmap [based on Moore’s law] plateaued, forcing a more fundamental question—one that, in retrospect, the entire industry will eventually have to confront,” the company said in a paper released on May 25.

Huawei’s Kirin chipset line will be the first to fully adopt the new law, and will be used in its flagship smartphone to be released this year. Huawei said it will use the law to improve the computing performance of its other AI chip products as well.

Before the US hit Huawei with strict export controls to block its use of advanced foreign technologies, the Chinese tech giant briefly became the world’s second largest smartphone maker by shipments, surpassing Apple and trailing only Samsung. Huawei’s smartphone business was hit particularly hard by the US crackdown. Whether the revamped Kirin chipset can meaningfully revive the company’s handset business remains to be seen.

Huawei said that with the latest chip design approach, it could match the 1.4-nanometer, also known as 14 angstrom, chip technologies of global chip leaders by 2031. This timeline would put the company just a few years behind TSMC and Intel, which are each planning for their own 1.4-nm chip production to start around 2028 and 2029.

Generally, the smaller the nanometer size, the more advanced the chips. However, it has become extremely challenging to pack more transistors into the same size of chip area and the semiconductor sector is exploring ways to go beyond Moore’s law.

While the upcoming Kirin chips are the first to fully adopt Her’s law, Huawei said it has already designed the new approach to varying degrees in designing 381 types of chips across consumer electronics, networking, and computing products over the past six years.

He said the company looks forward to working closely with scientists, engineers and industry partners around the world on the new law to keep driving the advancement of the chip and electronics industries.

“No single company can independently find all the answers along the path of semiconductor evolution,” she said.

With the chip industry as a whole wrestling with the slowing of Moore’s law, global chipmakers are stepping up efforts to expand and develop new chip packaging technologies by connecting and stacking different chips together to boost computing performance. AI chip leaders such as Nvidia, Google, and Amazon are exploring advanced packaging methods to more tightly connect and integrate GPUs (graphics processing units), CPUs (central processing units), and memory chips for faster and more efficient computing.

Huawei’s announcement comes as Jensen Huang, CEO of Nvidia, which is Huawei’s biggest rival in AI computing, recently said his company has basically conceded the Chinese market to Huawei amid the US export controls on AI chips to the country. AMD’s CEO Lisa Su also acknowledged that her company’s high-end AI chipset is unlikely to see meaningful sales in China.

This article first appeared on Nikkei Asia. It has been republished here as part of 36Kr’s ongoing partnership with Nikkei.