According to OFweek Robotics, China's National Development and Reform Commission (NDRC) has issued a clear warning to the country's booming robotics sector: focus on practical applications and sustainable growth rather than blind speculation. Speaking at an August 28 press conference, NDRC spokesperson Chao Li called for a disciplined approach to development, urging governments and tech firms to build on local resource endowments and industrial strengths rather than blindly following trends.

Nvidia reported second-quarter revenue for fiscal year 2027 of US$96.22 billion, a 106% year-over-year increase, driven by a 117% surge in data center revenue to US$89 billion. With third-quarter revenue of fiscal year 2027 projected to hit US$108 billion and full-year fiscal year 2028 revenue expected to grow around 70%, Nvidia revealed that current supply capabilities meet only about 70% of customer demand. This indicates that the primary constraint on Nvidia's sustained hyper-growth has officially shifted to the supply side.

Apple's Mac mini and Mac Studio are becoming unlikely fixtures in AI infrastructure. OpenAI has reportedly bought tens of thousands of the desktops for reinforcement learning and is seeking more, while Anthropic is renting Mac minis through Amazon Web Services, according to The Information. The systems are being used partly to train computer-use agents that operate software, test code and complete multi-step digital tasks.

Chip design company Agentrys announced last week that it raised US$24.5 million in funding from backers including MediaTek. It is one among an emerging crop of startups using agentic tools to power the design of semiconductors, a step in the chip process that still involves considerable amounts of time and labor.

US restrictions on China's access to advanced computing could expand beyond physical chip shipments to the remote use of computing resources, potentially increasing compliance requirements for cloud providers and data center operators in Southeast Asia.

MetaX posted its first half-year profit since listing, with rising GPU shipments and the mass production of its C600 accelerator strengthening its commercial position while China's domestic AI chip sector increasingly splits between profitable vendors and peers still absorbing heavy R&D costs.

OpenAI announced that it will end its contract allowing its models to be used on Cursor due to the latter's acquisition by SpaceX. The development indicates the tense personal relationship between AI leaders Sam Altman and Elon Musk months after an acrimonious lawsuit between the two and as both companies pursue the lucrative enterprise market.

On August 26, the Shanghai municipal government released its 15th Five-Year Plan for the Development of Strategic Emerging Industries, setting a target for added value in strategic emerging industries to reach CNY2.1 trillion (approx. US$312.5 billion) by 2030. The plan puts ICs, biopharmaceuticals, and artificial intelligence (AI) at the center of the city's industrial strategy, while shifting policy focus from expansion to higher-value areas such as advanced chips, advanced packaging, industrial software, and critical materials.

Asustek Computer (Asus) is accelerating its artificial intelligence (AI) transformation plan, announcing the rebranding of its AIoT Business Group into the Physical AI Solutions Business Group (PAS BG), with humanoid robots positioned as an important strategic market for the future. The move underscores the company's push beyond AI servers into agentic AI, edge AI, and physical AI under the chairman Jonney Shih's AI strategy.

As artificial intelligence (AI) technology rapidly evolves, physical AI for industrial applications is drawing growing attention because repeatable and scalable tasks play to AI's strengths. Deloitte Taiwan has just released a full report on physical AI ahead of a physical AI seminar it is organizing in Taichung on September 15.

A move by SpaceX to manufacture turbine blades in-house could ease a global bottleneck in AI infrastructure, where power equipment is now as constrained as chips. If successful, the plan could speed data center buildouts, shift supply chains, and strengthen Musk's control over the timing of new energy capacity.

India's semiconductor ambitions are expanding from packaging and testing to chip design, materials, mobile devices and data centers. New investments by CG Power, Infineon, Foxconn, Samsung and LG highlight rising industrial activity, while skills shortages, environmental constraints and import dependence remain challenges. Government incentives increasingly seek deeper domestic capabilities and higher value creation.

Universal Scientific Industrial (USI), part of ASE Technology Holding, is expanding its optical interconnect footprint for AI data centers. EugenLight Technologies, in which USI holds a controlling stake, has launched its ultra-high-power (UHP) external laser source form-factor pluggable (ELSFP) module series, targeting high-speed optical interconnect demand in next-generation AI compute clusters, high-performance computing (HPC), and high-density data centers, with mass production scheduled for the first quarter of 2027.

Tex Year is widening its specialty chemicals business beyond LCDs and electronics into AI servers, in-vehicle electronics, Mini LED, and optical lenses, while also targeting optical communications, silicon photonics (SiPh), and advanced packaging. The company says the segment is set to become its second growth engine as demand shifts toward higher-value applications.

Shinkong Synthetic Fibers Corporation (SSFC) is taking a cautiously optimistic view of operations in the second half of 2026. Despite geopolitical volatility and disruption to Middle East oil supply keeping international crude prices elevated in the short term, the company said its business should gain support from the US market's relative tariff advantage, gradual ramp-up of new higher-value products, recycled materials, AI applications, and the group's diversified investments.

AI infrastructure is entering a phase in which GPUs are no longer the only constraint. Nvidia's Vera Rubin generation is pushing rack power high enough to force parallel changes in cooling, power delivery and data center design.

As embodied intelligence takes physical form amid widespread production-capacity expansion in the humanoid robot sector, it is becoming increasingly possible to envision a society in which the boundaries between human intelligence and machine activity grow progressively blurred under AI's expanding reach.

As society ushers in the agentic AI era, the AI infrastructure space is simultaneously undergoing intense competition that increasingly tests the limits of technology, as industry players vie for top-tier compute.

As artificial intelligence (AI) chips continue to scale up in physical size, semiconductor competition is rapidly extending beyond front-end node scaling into advanced packaging. The South Korean semiconductor industry is eyeing glass substrates as its next major breakthrough, aiming to shore up its lagging packaging ecosystem and secure greater influence in AI accelerators, high-bandwidth memory (HBM), and system-on-chip integration markets.

China's humanoid robot industry is flourishing, and the recently concluded World Robot Conference (WRC) 2026 in Beijing revealed a notable shift in focus. Unlike previous years, when walking and motion capabilities dominated the competition, attention is increasingly moving toward robots' end-effectors.

Research Insight: Humanoid robots near mass production as suppliers face precision and ramp-up test

AI adoption continues to expand, but putting the technology to work and integrating it into everyday operations remain key hurdles. Speaking at the 2026 Taiwan Sustainability Summit, Yung-Hui Li, director of the Artificial Intelligence Research Center at the Hon Hai Research Institute, outlined Foxconn's approach to the generative AI era, covering the group's strategic direction, the technical path behind its in-house large language model FoxBrain, and the application of AI in smart manufacturing, energy conservation and carbon reduction.

The four major US cloud service providers (CSPs) have already rewritten their 2026 capital expenditure records, and ASIC vendors are increasingly confident that demand will remain strong through 2028 as customer spending and supply-chain visibility both extend farther out.

KKCompany Technologies, parent of KKBOX, will officially list on the Taipei Exchange's Emerging Stock Board on August 31, with chairman and CEO Steve Wang detailing, for the first time, the group's integration of PressPlay during a pre-listing investor briefing. The transaction, completed in the second quarter of 2026, is designed to expand KKCompany beyond music and AI multimedia into knowledge learning and the creator economy.