Largan Precision chairman En-ping Lin has questioned the viability of Micro LED-based optical links for AI data centers, citing concerns over power consumption, size, and transmission efficiency. Industry sources interviewed by DIGITIMES counter that while individual Micro LEDs transmit data more slowly than lasers, using hundreds of parallel channels could reduce overall system power consumption by simplifying the transmission architecture.
Microsoft's lease commitments that have not yet commenced reached US$329.1 billion in the fourth quarter of fiscal 2026, which ended in June, according to its latest regulatory filing. The total rose US$132.5 billion, or more than 67%, from the previous quarter, marking the company's largest quarterly increase on record.
Nvidia's share of AI chips in China's data center market has fallen by 34% in the first half of this year, according to a new report. The drop follows months of US restrictions on advanced chip sales to the country, Chinese authorities' push to rely less on Nvidia imports, and domestic players like Huawei moving to fill in this supply gap.
IC substrate maker Xinxing experienced a fatal workplace accident at its Yangmei plant on October 13, prompting the Taoyuan City Labor Inspection Office to order a partial shutdown in the affected area. The company said the incident has not affected production or operations so far and that it will cooperate fully with the investigation and implement preventive measures.
The debut of Google's Gemini 4 family marks a major milestone in frontier artificial intelligence, and Gemini 4 Argon is set to lead this next-generation launch. As autonomous AI capabilities scale rapidly, security incidents across the tech landscape highlight why robust containment controls have become the industry's top priority.
OpenAI is in early talks to acquire Trail of Bits, a cybersecurity research firm specializing in software vulnerabilities and open-source security, The Information reported on October 13, as the ChatGPT maker expands its push into enterprise cybersecurity.
Samsung Electronics and AMD have collaborated to validate Compute Express Link (CXL) memory technology designed to expand capacity and bandwidth for AI servers, as Samsung broadens its AI infrastructure strategy beyond high-bandwidth memory (HBM) to memory pooling and enterprise storage.
As AI cluster sizes grow, networking systems face challenges beyond higher bandwidth, including power consumption, cooling, cross-site coordination, and large-scale device operations. Cisco is expanding its AI data center technology strategy from high-bandwidth switch chips to cross-site networking, open network operating systems, and device management.
Cisco and Nvidia announced at the 2026 OCP Global Summit that they are jointly developing an ultra-dense small form factor pluggable (USFP) optical module designed to support 400Gb/s electrical signals per lane, targeting AI data centers' demand for higher bandwidth density, greater energy efficiency, and high-speed interconnects.
The AI accelerator market is set to reach US$1.4 trillion by 2030, up 150% from Bloomberg Intelligence's January 2026 estimate of US$560 billion, as higher prices—not just shipment growth—drive the market. Annual revenue is expected to surpass US$1 trillion in 2028, growing at a compound annual growth rate (CAGR) of 38% from 2026 to 2030.
CyberTAN Technology's sale of its Hsinchu Science Park factory and auxiliary facilities signals a broader shift toward higher-growth networking and AI infrastructure. The move may reshape how the Taiwanese company deploys capital, and it underscores continued global demand for data center, enterprise, and telecommunications equipment.
More coverage
