At the 2026 OCP APAC Summit, debate over networking architectures in cloud AI data centers sharpened as industry players said the main bottleneck is no longer compute; beyond memory, networking has become a key constraint on how far AI infrastructure can scale.
Univacco Technology posted record second-quarter and first-half 2026 revenue and profit, driven by its expanding global distribution network, an improved product mix and stronger demand in North America. The vacuum metallized film and precision coating maker also said July revenue stayed at a historically high level, setting a solid tone for the third quarter, which it described as the industry's peak season.
Compal chairman Ray Chen said the company was still a newcomer to AI servers two years ago but has now moved into the middle tier and wants to keep closing the gap with the leaders. However, opportunities to break into Nvidia and AMD's supply chains remain difficult to win, whereas enterprise OEM business offers a more promising opening for Compal to catch up with front-runners.
Nvidia CEO Jensen Huang has sought to reassure investors over the chipmaker's exposure to a newly announced initiative designed to mobilize more than US$500 billion in third-party capital for artificial intelligence infrastructure.
Naver has reportedly invested in US artificial intelligence developer Anthropic, adding another global AI company to a widening network that now spans frontier models, AI agents and large-scale computing infrastructure.
National Tsing Hua University (NTHU), one of Taiwan's leading research universities, has spent more than a decade cultivating a campus culture of professor-led startups — a track record that has already produced companies such as eMemory Technology, Heron Neutron Medical, Advanced Ceramic X (ACX), Apex Biotechnology (ApexBio), and Finesse Technology. Building on that history, the university announced on August 10 that it will expand the scale of its "NTHU Accelerator."
Global AI computing is rapidly moving beyond single-GPU systems toward AI Factory architectures, where agentic AI and inference workloads are pushing data centers to rely on networking, optics, packaging, cooling, power delivery and software co-design rather than raw accelerator counts. In a DIGITIMES interview, Nvidia Senior Vice President of Networking Gilad Shainer said the real unit of computing is now the entire data center, not an individual server.
