Nvidia cofounder and CEO Jensen Huang delivering the opening keynote speech of Computex 2025 in Taipei, Taiwan.
I-HWA CHENG/AFP via Getty Images
Amid an intensifying global race for cutting-edge AI technologies, billionaire Jensen Huang’s semiconductor design giant Nvidia and billionaire Terry Gou’s contract electronics maker Hon Hai Technology Group (Foxconn) announced they will partner with the Taiwanese government to build an “AI factory supercomputer.”
To develop the AI supercomputer, Nvidia will supply 10,000 Blackwell GPUs – the world’s most advanced chips for generative AI, part of the company’s next-generation GB300 systems – while Foxconn will provide AI infrastructure through its subsidiary, Big Innovation Company, as an Nvidia cloud partner. The resulting supercomputer aims to “significantly expand AI computing availability” for researchers and enterprises, including the Taiwan National Science and Technology Council and billionaire Morris Chang’s Taiwan Semiconductor Manufacturing Co. (TSMC).
“AI has ignited a new industrial revolution – science and industry will be transformed,” said Huang, cofounder and CEO at Nvidia, in a statement. “We are delighted to partner with Foxconn and Taiwan to help build Taiwan’s AI infrastructure, and to support TSMC and other leading companies to advance innovation in the age of AI and robotics.”
Often billed as the future of computing, AI supercomputers are powered by specialized hardware, such as GPUs and TPUs, consisting of more than hundreds of thousands of processing cores. Interconnected with high-speed networks, these processors can work in parallel to perform calculations on vast quantities of data. One prominent example is Hewlett Packard Enterprise’s (HPE) Frontier supercomputer, which can perform a quintillion (1,000,000,000,000,000,000) calculations per second – over a million times more than the number of stars in the Milky Way galaxy. While not explicitly designed for AI workloads, Frontier can help scientists complete complex AI-augmented tasks, such as modeling the lifespans of nuclear reactors or simulating the interactions of subatomic particles in elements.
“By harnessing advanced AI infrastructure, we empower our researchers to accelerate breakthroughs in semiconductor technology, enabling next-generation solutions for our customers and the world,” said C.C. Wei, chairman and CEO of TSMC, in a statement. “Leveraging this AI factory reinforces our commitment to pushing the limits of AI-driven innovation.”
The AI supercomputer collaboration comes as Foxconn, the largest supplier of Apple, has launched other initiatives with Nvidia while attempting to wean its reliance on Chinese manufacturing. Ahead of Taiwan’s annual Computex electronics trade show and conference, it announced the debut of its “Nvidia-accelerated” nursing robot, Nurabot, which aims to help hospital staff transport medication and samples in facilities across Taiwan. Last Wednesday, Foxconn received approval from India’s cabinet for a joint venture with billionaire Shiv Nadar’s IT hardware giant HCL Group to build a 37 billion rupees ($435 million) semiconductor plant in the northern state of Uttar Pradesh. Announced in a cabinet briefing, the new plant is slated to be operational by 2027.
Taiwan, home to several key players in the semiconductor supply chain, has long supplied most of the world’s semiconductors. Aside from TSMC, the world’s largest semiconductor foundry, other leading companies include billionaire Tsai Ming-kai’s Mediatek, which designs chips for both generative and edge AI, and billionaire Archie Hwang’s Hermes-Epitek, which specializes in optoelectronic equipment used to make advanced microprocessors.
MORE FROM FORBES
