The forums opened in Taipei on Monday and the exhibition halls open Wednesday, and by any measure this is the largest edition SEMI has run: more than 1,300 exhibitors, 4,300 booths, 18 national pavilions, over 100,000 people from 65 countries. Numbers like that are easy to skim past. The substance underneath them matters more, because the question dominating almost every session is not which accelerator is fastest. It’s whether the connections between accelerators can keep up at all.

China’s July customs data arrived on Friday with the headline pointing one way and the composition pointing somewhere considerably more interesting. Exports rose 23.9% year on year in dollar terms to $397.85 billion, beating the 22.2% consensus in the Reuters poll and the 22.88% figure from the Wind survey, but decelerating from June’s 27%. Imports climbed 27.5% to $285.35 billion, slower than June’s 36% jump and broadly in line with what economists expected. The surplus narrowed to $112.5 billion from $125.62 billion.

Taiwan’s technology sector has long derived strategic value from its position in semiconductor fabrication. What COMPUTEX 2026 makes visible is a second-order ambition: that Taiwan intends to compete not merely as a components supplier but as an end-to-end architect of AI infrastructure — from silicon to deployed operational systems.

GIGABYTE Technology’s showcase under the theme “Future Landing” is organized around a supply-chain logic rather than a product catalog. The company presents three operational states — Ready, Deployable, and Happening — that map to the full lifecycle of AI infrastructure: systems validated before shipment, modular clusters engineered for rapid field deployment, and AI actively running in production environments. The framing is deliberate. It positions GIGABYTE not as a hardware vendor but as an infrastructure integrator capable of compressing the timeline between procurement and operational readiness.

Edgecore Networks, a subsidiary of Accton Group and a leading provider of open networking solutions, has unveiled outcomes of the Innovative Optical and Wireless Network (IOWN) all-photonics network application at the “AI IMPACT: Smart Cities and Southern Taiwan Achievement Showcase,” held under the guidance of Taiwan’s National Science and Technology Council (NSTC).

The demonstration was jointly presented by the National Center for High-performance Computing (NCHC), Chunghwa Telecom (CHT), and Accton Technology / Edgecore Networks, with participation from Japan’s NTT and Taiwan NTT Systems. The collaboration applies IOWN’s All-Photonics Network (APN) and photonic-electronic convergence technologies to realize low-power “IOWN photonic computing” use cases — marking a significant milestone in Taiwan–Japan cooperation toward sovereign AI infrastructure.

China’s cloud infrastructure services market recorded $14.7 billion in spending during the fourth quarter of 2025, a 26% year-on-year increase and the third consecutive quarter of growth above 20%, according to research by Omdia. The figure is striking not merely as a commercial milestone but as a indicator of the pace at which the People’s Republic is building the computational substrate that underpins both its economic ambitions and its military modernization.

NVIDIA is reportedly adding Taiwan-based Nanya Technology as a new LPDDR5X memory supplier for its upcoming Vera Rubin AI platform, a notable supply-chain development that signals how aggressively the company is preparing for next-generation AI infrastructure demand. Nanya would become the first Taiwanese memory maker selected for this portion of the Rubin ecosystem, an area previously dominated by larger Korean and U.S. suppliers.

The strategic importance is less about one supplier and more about diversification. NVIDIA’s Vera CPU inside the Rubin platform uses high-capacity LPDDR5X memory, while Rubin GPUs are expected to rely on advanced HBM memory. That split memory architecture allows NVIDIA to spread sourcing across multiple vendors, reducing bottlenecks after the severe supply constraints seen during the AI boom.

TAITRA has confirmed that NVIDIA founder and CEO Jensen Huang will deliver the keynote address at COMPUTEX 2026, making official what had been widely anticipated: Taipei will once again serve as the opening frame for NVIDIA’s mid-year narrative push. The presentation takes place at the Taipei Music Center on June 1 at 11 a.m. Taiwan Time — Sunday, May 31 at 8 p.m. Pacific — and will be livestreamed globally through NVIDIA’s GTC Taipei keynote page, with a replay available afterward.