Taiwan's Keelung prosecutors detained Albatron Technology general manager Kevin Lu on Tuesday on suspicion of smuggling Supermicro AI servers to restricted markets, putting one of Taiwan's most active authorized distributors of American tech brands at the center of a US export control enforcement case.
Competition in China's humanoid robot market is driving down prices for dexterous hands and other key parts, with implications for suppliers and buyers worldwide. Rapid product cycles are forcing cost cuts, while technical barriers, especially in high-precision components, continue to shape which manufacturers can compete globally.
Amazon Web Services (AWS) is rolling out a new engineering division aimed at helping companies move beyond experimenting with artificial intelligence and start running it at the core of their operations.
South Korean AI chip designer Rebellions said on June 30 that it is acquiring AI inference optimization company SqueezeBits, as part of an effort to become a full-stack AI infrastructure provider rather than a chip designer alone.
Schneider Electric, the French energy management and automation giant, announced that it has agreed to acquire Cognite, a Norwegian industrial data and AI software company, in an all-cash deal valued at US$3.1 billion. The deal is meant to reinforce the former's software line-up as it positions itself for a future of AI-powered industrial automation.
Samsung Electro-Mechanics is in final-stage talks to supply multilayer ceramic capacitors (MLCCs) to a major US cloud provider for AI servers, while also expanding substrate production capacity and securing a separate silicon capacitor contract, according to Korean media reports and company statements.
Chief Telecom Inc. said a June 29 search by prosecutors and investigators over an alleged illegal smuggling case involving high-end AI servers bound for Hong Kong, Macau, and China has not materially affected its finances or operations. The case highlights growing global scrutiny of AI hardware supply chains and data center controls.
The AI infrastructure order boom is spilling into the second half of 2026, and server supply-chain players are turning increasingly upbeat about demand as Nvidia Vera Rubin, AMD Helios, AWS Trainium 3, and Google TPU all move into mass production.
A new partnership between Nvidia and Firmus aims to expand access to advanced AI computing for customers worldwide, including AI-native companies, enterprises, and independent software vendors. The deal underscores how demand for large-scale AI infrastructure is reshaping global technology markets, especially in the Asia-Pacific region.
Honda Motor has begun producing data-center batteries at an Ohio factory originally built to supply electric vehicles, as automakers and battery suppliers seek new uses for capacity while EV demand cools.
A name largely absent from the global supercomputing stage for years returned to the spotlight at ISC 2026 in Hamburg, Germany.
Nvidia does not make robots, but it is becoming a key force behind embodied intelligence companies. At Automate 2026, North America's largest industrial automation show, on June 22, Deepu Talla, vice president of Nvidia's robotics business, said on stage that the company hired 18,600 man-years of engineers to bring the safety architecture of autonomous driving to robots.
Kunlunxin, the semiconductor subsidiary of Chinese search engine giant Baidu, is targeting a US$50 billion valuation for its Hong Kong public offering. The company is also asking investors to commit to buying its chips as a condition of participation, according to The Information, underscoring the competitive dynamics shaping chip makers as Beijing moves to strengthen its domestic AI supply chain.
Kyocera is committing JPY650 billion (approx. US$4 billion) to its components businesses through fiscal 2031, as AI data center investment and semiconductor equipment spending lift demand for ceramic parts, optical communication packages, and advanced semiconductor packaging materials.
Wistron is stepping up factory spending in the US, Taiwan, and Southeast Asia to meet rising demand for AI servers. The expansion signals how global supply chains are shifting to support faster deployment of AI hardware, with California emerging as a key hub for its customers globally.
With growing demand for AI server cooling and power management solutions, power semiconductor design company Potens reported that revenue from its server-related business has risen from 4.5% of total revenue in 2025 to 13.5%, a significant jump that reflects strong momentum in the segment. The company also remains optimistic about continued expansion in the AI, automotive, and motor control markets. Order transfers from Western manufacturers seeking to reduce reliance on China are also materializing.
As Physical AI moves closer to commercial reality, its global impact may depend less on humanoid robots that can run or dance and more on their ability to safely grasp, lift, and manipulate objects. A Massachusetts Institute of Technology professor said the central challenge is touch, force, and real-world control, not simply bigger models or faster computing.
