Information/Technology - 3. page

Yang Xuezhi, a former Huawei scientist and telecommunications expert known for inventing the 4G Soft Frequency Reuse (SFR) technology, recently sparked controversy after publishing an article criticizing Huawei’s newly announced “Tao (τ) Law.”

The debate began after Huawei Semiconductor Business President He Tingbo introduced the “Tao Law” on May 25, 2026. Huawei describes it as a chip-design concept that improves semiconductor performance through circuit architecture optimization, signal-timing improvements, and system-level efficiency gains, rather than relying primarily on transistor miniaturization as envisioned by Moore’s Law.

Yang argued that the technology represents a valuable engineering optimization rather than a fundamental scientific breakthrough. He contended that the concepts underlying the “Tao Law,” such as reducing signal delays and optimizing circuit timing, are well-established industry practices and do not justify being labeled a new “law.” He accused Huawei of repackaging existing technologies, exaggerating their significance through marketing, and engaging in what he described as a long-standing pattern of overstating research achievements.

The article was subsequently removed from Chinese online platforms. In response, a number of Chinese commentators and technology writers published rebuttals defending Huawei’s technological contributions. Others attacked Yang personally, labeling him a “loser” and arguing that his criticism stemmed from resentment over his departure from Huawei in 2012 rather than from objective academic concerns.

Sources:
1. Aboluo, June 14, 2026
https://www.aboluowang.com/2026/0614/2395705.html
2. Guancha.com, May 29, 2026
https://user.guancha.cn/main/content?id=1661107&page=1
3. Sina, June 3, 2026
https://www.sina.cn/news/detail/5305769256815587.html

Huanqiu Times reported that a recent IDC study showed China’s AI-enabled Enterprise Resource Planning (ERP) market reached US$315.7 million in 2025, representing year-on-year growth of 96.1 percent.

IDC noted that the figure reflects enterprises’ additional investment in AI capabilities during ERP upgrades rather than the overall ERP software market. According to the report, the rapid growth signals a fundamental transformation in ERP systems rather than the simple addition of AI functions.

Companies are no longer satisfied with using ERP systems merely to record business activities. Instead, they increasingly expect ERP platforms to participate directly in business execution, shifting from a model in which users search for functions to one in which systems can understand user intent and proactively provide services.

Source: Huanqiu Times, June 5, 2026
https://tech.huanqiu.com/article/4Rr63uq5wM1

Xinhua News Agency reported that Chinese researchers announced the successful development of “Jiuzhang-4,” a programmable quantum computing prototype featuring 1,024 squeezed-state quantum inputs and 8,176 modes, which Chinese scientists describe as a new world record in photonic quantum information technology.

According to the research team, “Jiuzhang-4” can manipulate and detect quantum states involving up to 3,050 photons, significantly surpassing the 255-photon capability of the earlier “Jiuzhang-3” system. The team stated that the prototype can solve Gaussian boson sampling problems at a speed 10^54 times faster than the world’s most powerful conventional supercomputers.

University of Science and Technology of China professor Lu Chaoyang said the team developed a high-efficiency optical parametric oscillator light source and a spatiotemporal hybrid-coded interferometer to support the system’s performance.

The report also highlighted China’s broader advances in quantum computing competition. After Google announced a 53-qubit quantum computer in 2019, China introduced both the 76-photon “Jiuzhang” photonic quantum prototype and the 56-qubit superconducting prototype “Zuchongzhi-2,” which Chinese researchers say made China the only country to achieve “quantum computational advantage” in both photonic and superconducting approaches. The report further stated that Chinese researchers reached a milestone in quantum error correction by the end of 2025, achieving “below-threshold” operation in which error correction improves as additional corrections are applied.

Source: Xinhua, May 13, 2026
https://www.news.cn/tech/20260513/8d58ad5254d0483f8fbe5cbeb108085b/c.html

Major Taiwanese news network Liberty Times Network (LTN) recently reported that, NVIDIA is expected to release its first Windows PC processor. This marks the AI chip giant’s official entry into the PC market and provides Microsoft with a new opportunity to revitalize its AI-centric computer strategy.

NVIDIA posted the coordinates “25.0528, 121.5990” on the social media platform X, referring to the Taipei Music Center (PKC) campus, where CEO Jensen Huang will a speech on June 1st. Microsoft and ARM subsequently followed suit by posting the same message, hinting at a new era that is about to begin in the PC industry. Microsoft’s initial foray into the AI PC market suffered a setback, but the addition of NVIDIA may give it a second chance. This time, it has the support of the world’s most popular chip manufacturer. NVIDIA and Microsoft plan to jointly showcase these new systems at Computex and the Microsoft Build developer conference in San Francisco.

Sources revealed that the PCs being launched are expected to include Microsoft’s Surface series and related models from Dell. This launch will also be NVIDIA’s most significant expansion to date, extending its dominance in AI data center chips into the mass-market consumer computer market. Microsoft and Dell declined to comment, NVIDIA did not respond to requests for comment.

Source: LTN, May 31, 2026
https://ec.ltn.com.tw/article/breakingnews/5455351

An independent cybersecurity researcher using the alias NetAskari recently stumbled upon an unsecured Chinese police surveillance dashboard — and found his own passport photo, phone number, and high-speed rail seat history staring back at him.

The exposed system was a demo panel built for the public security bureau in Zhangjiakou, Hebei province, a host city for the 2022 Winter Olympics. Though not yet fully connected to live data streams, the panel contained real datasets that revealed the ambitions of China’s evolving surveillance apparatus: a shift from standalone street cameras toward a predictive, always-on social control network.

The system’s reach is striking. Beyond standard China Central Telection (CCTV) feeds, it logs the exact train car and seat number when a target arrives by high-speed rail, captures facial recognition scans from ski resort turnstiles, and tracks fuel consumption, shopping locations, and visits to areas flagged as “frequent petitioning zones.” The goal, NetAskari noted, is to stitch together physical movements, spending habits, and digital footprints into a seamless “holographic dossier” — an approach comparable to U.S. data analytics firm Palantir.

Foreign nationals, particularly journalists and citizens from Five Eyes countries (the U.S., UK, Australia, New Zealand, and Canada), receive disproportionate attention. Some foreign correspondents are tagged for real-time tracking, triggering automatic police alerts the moment they enter a jurisdiction. This effectively renders covert on-the-ground reporting obsolete — authorities can anticipate a journalist’s itinerary via payment records and ticket purchases, and quietly pressure sources before any interview takes place.

The system also auto-generates social network maps from surveillance footage, visualizing who associates with whom and for how long. In 2025, Shanghai’s Putuo district police reportedly tendered a contract worth approximately $200,000 (around 1.45 million yuan) for a comparable “holistic personnel profile system.”

As NetAskari concluded in his report, within this infrastructure, people are reduced to data points — patterns to be monitored, predicted, and controlled.

Source: Deutsche Welle, May 20, 2026
https://p.dw.com/p/5E12T

Major Taiwanese news network Liberty Times Network (LTN) recently reported that, Nvidia CEO Jensen Huang revealed during a visit to Taiwan on Saturday (May 23, 2026) that Nvidia is entering a brand new CPU market with a potential scale of US$200 billion, which will allow NVIDIA to reach a new peak of dominance in artificial intelligence.

Few companies receive as much attention as Nvidia. As one of the most important players in the artificial intelligence (AI) ecosystem, Nvidia’s role is often seen as a benchmark for the entire AI field. Jensen Huang’s unexpected statement will propel Nvidia from a graphics chip company into an AI hardware giant, and ultimately into a comprehensive AI application ecosystem company. Huang pointed out that Nvidia’s upcoming Vera Rubin platform includes a central processing unit (CPU) specifically designed for intelligent agents and artificial intelligence. This transformation also opens up new revenue streams for Nvidia and could be a key step in the company’s development.

Nvidia’s latest quarterly revenue reached US$81.6 billion, an 85 percent year-over-year increase. More than 92 percent of this revenue came from its data center business. In the foreseeable future, data center hardware will remain Nvidia’s primary revenue source, but the launch of the Vera Rubin platform positions the company to become one of the world’s leading CPU suppliers.

Source: LTN, May 23, 2026
https://ec.ltn.com.tw/article/breakingnews/5447290

Primary Taiwanese news agency Central News Agency (CNA) recently reported that, Google and Blackstone Group plan to jointly establish an artificial intelligence (AI) cloud company using Google’s proprietary chips. Blackstone Group, the world’s largest alternative asset management firm, is expected to invest US$5 billion in equity and hold a majority stake in the unnamed startup.

Sources familiar with the matter revealed that Google will provide the new company with software, services, and hardware, including a dedicated chip known as a Tensor Processing Unit (TPU). Google senior executive Benjamin Treynor Sloss will become the new company’s CEO.

This is Google’s largest attempt to date to sell its self-developed chips and commercialize them. This move will intensify the direct head-to-head competition between Google and Nvidia, the dominant player in the AI computing market.

As the computing power required to train and run advanced AI models reaches unprecedented levels, Google and Blackstone Group decided to join forces. Currently, most large AI companies rely on the computing infrastructure built around Nvidia-based chips, at least to some extent.

Source: CNA, May 19, 2026
https://www.cna.com.tw/news/ait/202605190035.aspx