In a bold move that could reshape the AI landscape, Nvidia has announced it will begin selling its advanced chip-to-chip communication technology—NVLink Fusion—to other companies. The announcement was made by CEO Jensen Huang during his keynote at the Computex AI event in Taipei on May 19, 2025.
This powerful innovation, originally designed for Nvidia’s own high-performance systems, is now being shared with the wider tech industry. The goal? To help other chipmakers build faster, more efficient custom AI systems by enabling seamless communication between multiple chips.
What is NVLink Fusion?
NVLink Fusion is Nvidia’s upgraded interconnect technology that allows chips to transfer massive volumes of data between each other at lightning speed. It’s essential for advanced AI models that require multiple processors working in unison. MediaTek and Marvell Technology have already committed to adopting the new NVLink Fusion in their future custom chip projects—signaling strong industry interest.
This move marks a major milestone in Nvidia’s evolution. Once known mainly for making graphics chips for gamers, the company now dominates the AI hardware space. Its chips are at the heart of today’s most powerful AI systems, including those running models like ChatGPT.
Nvidia’s Vision for AI Expansion
During the event, Huang didn’t just unveil new tech—he shared Nvidia’s broader vision. He revealed plans to open a new Taiwan headquarters in Taipei’s northern suburbs, strengthening the company’s presence in one of Asia’s key tech hubs.
Huang’s keynote also looked back at how Nvidia has transformed over the years. “There was a time when 90% of our presentations were about graphics chips,” he said. “Now, it’s all about AI.”
As AI technology matures, the focus is shifting from simply training large models to running them in practical applications—and Nvidia is determined to lead both fronts. In March, the company launched its Blackwell Ultra chips, and teased upcoming Rubin and Feynman processors, expected in 2028.
AI for Desktops Too
In addition to NVLink Fusion, Nvidia introduced DGX Spark—a desktop version of its AI computer designed for researchers. Huang confirmed the system is already in full production and will be available within weeks.
Why This Matters
The launch of NVLink Fusion for external use is a strategic play that opens the door for other tech firms to accelerate their AI hardware development. By creating an open ecosystem around its powerful chip-linking tech, Nvidia is setting the stage for a new wave of innovation in AI infrastructure.
As the AI arms race continues, Nvidia isn’t just leading the charge—they’re enabling others to run beside them.
