TAIPEI: Nvidia boss Jensen Huang announced plans for Taiwan’s “first AI supercomputer”, as he showcased on Monday (May 19) the company’s latest advances in artificial intelligence.
Global semiconductor chip giants have gathered in Taiwan for the island’s top tech expo, Computex, as the sector grapples with the impact of US tariffs and disrupted supply chains.
Huang said Nvidia would work with Taiwanese tech giants Foxconn and TSMC as well as the government to build Taiwan’s “first AI supercomputer here for the AI infrastructure and AI ecosystem”.
“Taiwan doesn’t just build supercomputers for the world … we’re also building AI for Taiwan,” Huang said in a keynote address, describing the island as the “centre” of the industry.
“Having a world-class AI infrastructure in Taiwan is really important.”
Taiwan-born Huang also spotlighted a further upgrade to Nvidia’s Blackwell processing platform, as well as new hardware and software for robotics and “AI agents” that can perform company tasks.
And he announced a new version of Nvidia’s NVLink technology, enabling customers to build semi-custom AI infrastructure.
“In 10 years time, you will look back and you will realise that AI has now integrated into everything and in fact we need AI everywhere,” Huang said, wearing his trademark black leather jacket.
Computex will draw computer and chip companies from around the world to Taiwan, whose semiconductor industry is critical to the production of everything from iPhones to the servers that run ChatGPT.
Taiwan produces the bulk of the world’s most advanced chips, including those needed for the most powerful AI applications and research.
Qualcomm CEO Cristiano Amon announced the company planned to expand into data centres, but he did not elaborate.
Top executives from MediaTek and Foxconn will also speak at Computex, where advances in moving AI from data centres into laptops, robots and cars are in the spotlight.
Tech expert Paul Yu told AFP the industry was at a “critical juncture” for AI hardware development.
“Over the past two and a half years, significant investment has driven rapid advances in AI technology,” said Yu, of Witology Markettrend Research Institute.
He added that “2025 to 2026 will be the crucial period for transitioning AI model training into profitable applications”.
Read the full article here
