Huawei Bets on Speed Over Shrinking Transistors to Sidestep US Chip Sanctions

 A logo for Huawei is seen during the KubeCon + CloudNativeCon Europe hosted by the Cloud Native Computing Foundation (CNCF) in Paris, France, March 20, 2024. (Reuters)
A logo for Huawei is seen during the KubeCon + CloudNativeCon Europe hosted by the Cloud Native Computing Foundation (CNCF) in Paris, France, March 20, 2024. (Reuters)
TT

Huawei Bets on Speed Over Shrinking Transistors to Sidestep US Chip Sanctions

 A logo for Huawei is seen during the KubeCon + CloudNativeCon Europe hosted by the Cloud Native Computing Foundation (CNCF) in Paris, France, March 20, 2024. (Reuters)
A logo for Huawei is seen during the KubeCon + CloudNativeCon Europe hosted by the Cloud Native Computing Foundation (CNCF) in Paris, France, March 20, 2024. (Reuters)

Huawei's new chip design principle focused on boosting transmission speed rather than continuing to shrink semiconductors offers a path for China to build cutting-edge chips despite US sanctions, though whether it represents a true breakthrough remains to be seen.

China has been barred since 2019 from importing ASML's most advanced extreme ultraviolet (EUV) lithography machines, curbing the ability of its chipmakers to keep up with global leaders like Taiwan's TSMC in relying on ever-smaller manufacturing processes that make chips more powerful.

For decades, the semiconductor industry has been governed by Moore's Law - the observation that the number of transistors on a microchip doubles roughly every two years.

Huawei this week unveiled an alternative approach: cutting the time signals take to move through chips and larger computing systems using a principle it calls the Tau Scaling Law.

Its central technique, LogicFolding, aims to arrange logic, analogue and memory circuits in stacked, more tightly connected structures, potentially improving density, efficiency and clock speeds over the next decade.

Proponents see it as ‌a way to ‌extend chip progress as manufacturing advances begin to slow.

"For Huawei, chips face two key constraints. ‌One ⁠is inevitable that Moore's ⁠Law will hit a physical 'wall' within the next decade," He Tingbo, the president of Huawei's semiconductor business, told China's People's Daily this week.

"The other is accidental because of the external restrictions that Huawei encountered this 'wall' earlier than its peers," she said, in a likely reference to US sanctions on importing advanced EUV machines.

But others argue that reducing latency has always been part of semiconductor design and that many of the underlying ideas resemble existing work in three-dimensional (3D) stacking, advanced packaging and system optimization.

"This is a breakthrough for Huawei, but it's not a threat for TSMC," Nvidia CEO Jensen Huang told reporters in Taipei on Thursday. "TSMC has been using die stacking and 3D packaging for how long now? Almost ⁠10 years. And so TSMC's technology is very advanced."

NOT A NEW CONCEPT?

In the race to ‌build more powerful computing systems, the chip industry has already embraced advanced packaging technologies ‌that stack chips vertically.

TSMC has been at the forefront with its packaging technology called SoIC, which enables more tightly integrated heterogeneous chiplets to reduce ‌size and improve performance.

Memory chip makers such as SK Hynix and Samsung Electronics also use advanced 3D stacking and packaging technologies ‌to produce multi-layer memory chips, a key component of AI chipsets, and to improve power efficiency and performance.

Huawei believes LogicFolding may actually go beyond the techniques commonly used in 3D integrated circuit stacking, thanks to "very finely and carefully split the critical paths of logic circuits across multiple layers," according to Liao Heng, chief scientist at Huawei Semiconductor.

But Bernstein analysts cautioned in a note that while stacking multiple chip layers boosts transistor density, it also increases power ‌density and risks overheating chips. Production yields and costs will be another barrier for adoption, they added.

Huawei's own roadmap also points to those challenges. Huawei's He said the approach would require ⁠new semiconductor design tools suited to ⁠folded chip architectures, as well as better ways to manage heat across devices ranging from smartphones to large AI data centers.

"With the methodology of not optimizing the area on a chip level, but on a system level based on time, that will dramatically change the capability requirements for the EDA (electronic design automation) vendors," said Handel H. Jones, CEO of International Business Strategies, during a panel discussion on Tau Scaling on Tuesday.

Mainstream EDA software produced by vendors like Cadence Design Systems and Synopsys plays a crucial role in creating blueprints for sophisticated semiconductor devices.

EYES ON NEW KIRIN CHIP

Huawei's most concrete claims centered on a new Kirin smartphone chip that will be launched later this year, which would be the first to use its LogicFolding architecture.

Compared with its earlier single-layer design, the new chip would improve power efficiency by 41%, and raise the chip's peak operating speed by nearly 13%, Huawei's He said in a speech on Monday.

Those figures would be significant if achieved at commercial scale. But Huawei did not provide production yield information, cost comparisons or a clear explanation of how the gains would compare with rival chips made using more advanced process nodes.

"There's nothing concrete that can be independently verified or benchmarked against other players at the moment," said Lian Jye Su, chief analyst at tech research firm Omdia.



China Smartphone Shipments Fall for 6th straight quarter - IDC

FILE PHOTO: The Huawei logo at the World Artificial Intelligence Conference (WAIC) in Shanghai, China, July 17, 2026. REUTERS/Go Nakamura/File Photo
FILE PHOTO: The Huawei logo at the World Artificial Intelligence Conference (WAIC) in Shanghai, China, July 17, 2026. REUTERS/Go Nakamura/File Photo
TT

China Smartphone Shipments Fall for 6th straight quarter - IDC

FILE PHOTO: The Huawei logo at the World Artificial Intelligence Conference (WAIC) in Shanghai, China, July 17, 2026. REUTERS/Go Nakamura/File Photo
FILE PHOTO: The Huawei logo at the World Artificial Intelligence Conference (WAIC) in Shanghai, China, July 17, 2026. REUTERS/Go Nakamura/File Photo

China's third-quarter smartphone shipments fell 15.4% from a year earlier to 57.9 million units, research firm IDC said on Tuesday.

Huawei was the only major vendor to post double-digit growth, Reuters reported.

The firm shipped 12.2 million units and led the market with a 21% share, IDC data showed.

Apple ⁠was second with 17.8%, ⁠though its shipments in China fell 4.3%.

Chinese rivals Oppo, Vivo and Xiaomi saw shipment declines of 17.6%, 24.9% and 20.4%, respectively.

The decline comes amid weak consumer spending in ⁠China. Over the past few quarters, phone makers have also been hit by rising prices for memory chips and other components.

Vendors have been passing higher costs on to consumers, pushing up retail prices even as shipments shrink, said Arthur Guo, a senior analyst at IDC.

IDC said Huawei's broad product lineup and relatively ⁠modest price ⁠increases helped it extend its run this year. Apple's new iPhones sold better than expected, but having fewer models this year weighed on its shipments.

Honor was the only leading brand whose shipments grew from the previous quarter, which IDC said suggested a possible rebound.


Anthropic AI Model Submits False Homicide Tip to Police Website

The Anthropic logo is seen in this illustration taken May 20, 2024. (Reuters)
The Anthropic logo is seen in this illustration taken May 20, 2024. (Reuters)
TT

Anthropic AI Model Submits False Homicide Tip to Police Website

The Anthropic logo is seen in this illustration taken May 20, 2024. (Reuters)
The Anthropic logo is seen in this illustration taken May 20, 2024. (Reuters)

An Anthropic artificial intelligence model submitted a false homicide tip through a Philadelphia police website, the department said in a statement, the latest incident of rogue behavior from the powerful technology. 

Anthropic notified authorities earlier this week and said that an automated testing process was responsible for the submission, ‌according to the ‌police. 

The fast-advancing technology has become a matter ‌of ⁠keen national interest amid ⁠reports of AI agents hacking into corporate computer networks, and warnings from researchers that it could one day pose an existential threat to humanity. In September, Anthropic rival OpenAI apologized for the hacking of an Australian health data portal by a rogue AI agent, the first known instance of an AI agent exploiting a government website. 

Previous ⁠incidents have involved AI agents hacking into vulnerable systems ‌or commandeering unsanctioned platforms to communicate ‌with one another. This is the first known case in which a rogue ‌AI appears to have tried to communicate a bogus tip ‌to authorities. 

The Philadelphia police department said the tip "was flagged as spam and was never forwarded to the Real-Time Crime Center for investigative vetting or dissemination," and added that Anthropic had told them the company intended to publish ‌a report on the incident on Friday. 

Anthropic did not immediately respond to a request for comment. 

Pennsylvania law ⁠specifies it ⁠is generally a misdemeanor crime for "a person" to knowingly give false reports to law enforcement authorities. This includes "information relating to an offense or incident when he knows he has no information relating to such offense or incident." 

Police quoted Anthropic as telling them the testing process was stopped after the incident was uncovered. 

The false tip was submitted through PhillyUnsolvedMurders.com concerning an unsolved homicide. The tip, dated July 18, 2026, purported to come from someone who might have information about the case, the police said. 

Police said they did not have evidence of unauthorized access to their systems or compromise of police department data. 


OpenAI Says it Has Fired 3 Researchers for Violating Sensitive Information Policy

FILE PHOTO: A keyboard is placed in front of a displayed OpenAI logo in this illustration taken February 21, 2023. REUTERS/Dado Ruvic/Illustration/File Photo
FILE PHOTO: A keyboard is placed in front of a displayed OpenAI logo in this illustration taken February 21, 2023. REUTERS/Dado Ruvic/Illustration/File Photo
TT

OpenAI Says it Has Fired 3 Researchers for Violating Sensitive Information Policy

FILE PHOTO: A keyboard is placed in front of a displayed OpenAI logo in this illustration taken February 21, 2023. REUTERS/Dado Ruvic/Illustration/File Photo
FILE PHOTO: A keyboard is placed in front of a displayed OpenAI logo in this illustration taken February 21, 2023. REUTERS/Dado Ruvic/Illustration/File Photo

OpenAI fired three of its researchers last week after an investigation found they violated policies on handling sensitive information, the AI major said in a post on X on Friday.

OpenAI's statement comes after the dismissed researchers Jasmine Wang, Tomek Korbak and Mikita Balesni published a ⁠letter outlining the circumstances ⁠of their termination, arguing that their abrupt firing could create uncertainty among remaining employees and undermine the culture that previously allowed researchers to raise safety concerns.

"I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation," Balesni said in an X post that accompanied the letter.

OpenAI, ⁠however, asserted that the dismissal was not related to raising safety concerns or speaking out.

"Our internal investigation uncovered a significant breach of trust beyond what's outlined in the letter they published and we stand by the decision to not continue their employment," Reuters quoted OpenAI as saying.

"Safety and research debates happen every day at OpenAI, often spirited and highly critical. We actively encourage these discussions and consider them essential to making the right decisions...We have not and do not terminate any of our employees for raising ⁠concerns," it ⁠said.

While OpenAI did not reveal the specifics of the alleged violations, Wang, Korbak, and Balesni said in their letter that they were not the source of a news article published on The Information last month about security concerns around OpenAI's latest AI model 'Astra'.

This comes as current and former researchers at major AI firms OpenAI, Google DeepMind, and Anthropic warn that companies are doing too little to guard against the potential fallout of building self-improving AI systems that could become difficult for humans to control.

In July, OpenAI agents broke out of their testing arena and hacked AI firm Hugging Face.