AI servers and advanced packaging continue to drive demand for semiconductor manufacturing, outsourced semiconductor assembly and test (OSAT), and equipment. Machvision said Advanced Semiconductor Engineering's (ASE's) expansion is helping make semiconductors and substrates its fastest-growing product lines.

Under Lisa Su's 12-year tenure at Advanced Micro Devices (AMD), the company has gradually moved beyond its role as Intel's challenger in the CPU market. But in the accelerator market for training and running AI models, old rival Nvidia still holds the upper hand.

Facing a prolonged memory supply crunch, Tesla CEO Elon Musk has announced steep cuts to the memory specifications of the company's AI5 and AI6 chips to secure enough memory for Optimus mass production. Musk said the downgrades will not affect the humanoid robot's performance.

A China-backed leasing firm appears to have helped Chinese companies buy restricted Nvidia Blackwell platform chips through financial leasing, according to a recent report. The transactions involved 32 Asus servers equipped with B300 chips, highlighting how US export controls face circumvention risks through finance and supply-chain channels.

Chenming Electronic Tech (UNEEC) president Charles Lo said the fourth quarter is a traditional peak season and that new models such as VR200 will enter mass production in the fourth quarter, helping revenue recover toward the company's former monthly level of more than NT$1 billion (US$31.3 million). For January-August 2026, revenue reached NT$7.9 billion, with L11 system integration accounting for 18%.

TSMC is reportedly weighing a new manufacturing campus in Texas that could add tens of billions of dollars to its US expansion, potentially creating a second major American production hub alongside Arizona as AI chip demand continues to strain advanced-node capacity.

To address chip talent shortages, Taiwan has a law in place to promote industry-academia collaboration by establishing semiconductor schools. However, the current financing scheme jointly supported by the National Development Fund (NDF) and contracted companies is deemed insufficient to meet the needs of the semiconductor schools, so the National Science and Technology Council (NSTC) has now promised extra subsidies.

JCET Group signed an agreement on September 29 for a high-end advanced packaging and testing project in Wuxi, Jiangsu Province. Wuxi municipal leadership pledged operational support to help JCET scale its manufacturing footprint, promoting regional integration and supporting China's semiconductor ecosystem, according to Sina Finance.

Data center AI accelerator package power has climbed from about 500W in the A100 generation to roughly 3,500–4,000W for Rubin Ultra-class products expected in 2027, pushing supply current into the thousands of amperes. In conventional frontside power delivery, current must cross the full metal interconnect stack before reaching the transistors, while power and signal lines compete for routing resources. Intel Foundry estimates that raising a single compute package from 1kW to 5kW increases power delivery network I²R losses from 89W to 2,222W, cutting the share of input power reaching the transistors from 66.0% to 41.8%.

AI-driven memory demand is surging, and Micron has recently accelerated capacity expansion across multiple regions worldwide, including Taiwan, the US, Japan, Singapore, and India. Donghui Lu, vice president of Micron's DRAM Asia frontend operations and head of Micron Taiwan, said the company's capex for construction over the next few years will be enormous, ranking it among the world's biggest construction investors, and stressed the keyword "together" as it works with Taiwanese equipment and materials suppliers to enter overseas markets.

Micron is using an externally manufactured base die in NVHBM, a custom high-bandwidth memory (HBM) product it is co-developing with Nvidia, as the company bets that customization can unlock added value even when manufacturing is outsourced. The move underscores how HBM competition is expanding beyond memory stacks into logic circuitry and system integration.