At the Computex AI exhibition in Taipei, Nvidia CEO Jensen Huang announced bold steps to reinforce the company’s dominance in the rapidly evolving artificial intelligence chip market. Central to the announcement was the launch of NVLink Fusion, an enhanced version of Nvidia’s chip interconnect technology that will now be licensed to other chip designers. This move aims to accelerate the development of powerful custom AI systems by enabling faster chip-to-chip communication.
NVLink, originally developed by Nvidia years ago, has been a key technology in enabling data exchange between multiple processors, such as in Nvidia’s flagship GB200, which combines graphics and CPU units for superior AI performance. With NVLink Fusion, Nvidia is expanding its influence beyond its own hardware ecosystem, empowering a broader range of AI innovators to build scalable, multi-chip architectures.
In addition to the new technology, Huang revealed plans to establish a dedicated Nvidia headquarters in northern Taipei, underscoring the company’s commitment to deepening ties with Taiwan, a critical hub in the global semiconductor supply chain.
Reflecting on Nvidia’s evolution, Huang noted that the company has transformed from a video game graphics chip maker to the leading force behind the AI revolution ignited by innovations like ChatGPT. He highlighted how Nvidia’s focus has shifted from graphics to AI chips and systems, supported by a robust software stack.
Further cementing its AI ambitions, Nvidia is set to release several new chip generations, including the Blackwell Ultra later this year, with follow-up processors like Feynman scheduled for 2028. Huang also introduced DGX Spark, a desktop AI computing platform aimed at researchers, which is entering full production and expected to ship imminently.
The Computex event, hosting around 1,400 exhibitors, marks a crucial gathering for the industry, especially amid ongoing geopolitical tensions and U.S. tariff threats designed to reshape chip manufacturing locations.
Nvidia’s announcements signal its intent not just to lead in AI hardware innovation, but to influence the wider chip ecosystem and global semiconductor landscape.
