In a groundbreaking announcement at Computex 2025 in Taipei, Nvidia introduced NVLink Fusion, a cutting-edge technology designed to enhance chip-to-chip communication for artificial intelligence systems.
Unveiled by CEO Jensen Huang during a keynote at the Taipei Music Center on May 19, 2025, NVLink Fusion allows chip designers to build powerful, custom AI systems by linking multiple chips for seamless data exchange.
This innovation addresses the growing demand for efficient data transfer in AI applications, positioning Nvidia as a leader in the AI hardware ecosystem.
The Computex exhibition, running from May 20 to 23, showcases this technology amid 1,400 exhibitors, which marks a turning point for the global tech industry.
NVLink Fusion: A Leap in AI System Design
NVLink Fusion builds on Nvidia’s earlier NVLink technology, originally developed to handle massive data exchanges in systems like the GB200, which integrates two Blackwell GPUs with a Grace processor.
Unlike its predecessor, NVLink Fusion is now available for licensing to other chip designers, enabling them to create high-performance AI systems.
Industry giants Marvell Technology and MediaTek have already committed to adopting NVLink Fusion for their custom chip projects, signaling strong market confidence in Nvidia’s innovation.
This move broadens Nvidia’s influence beyond its traditional role as a GPU manufacturer, reinforcing its pivot to comprehensive AI solutions.
ALSO READ | AMD’s $4.9B ZT Systems Acquisition Fuels AI Data Center Ambitions.
Desktop AI PCs and Taiwan Expansion
Huang also announced that Nvidia’s DGX Spark, a desktop AI chip tailored for researchers, is now in full production and will be available within weeks. This development democratizes access to high-performance AI computing for individual researchers.
Additionally, Nvidia revealed plans to establish a new headquarters in Taipei’s northern suburbs, underscoring its commitment to Taiwan’s tech ecosystem.
This expansion aligns with Taiwan’s role as a global hub for semiconductor innovation, especially as geopolitical tensions and U.S. tariff threats push companies to diversify manufacturing.
Did You Know?
Nvidia’s market capitalization surpassed $3 trillion in 2024, making it one of the world’s most valuable companies, driven by its AI chip dominance.
Nvidia’s Evolution and Industry Impact
Once known primarily for video game graphics chips, Nvidia has transformed into the dominant force in AI chip production since the AI boom sparked by ChatGPT in 2022. Huang highlighted this shift during his keynote, noting that his presentations now focus almost entirely on AI technologies.
Nvidia’s recent developments include CPUs designed for Microsoft’s Windows using Arm Holdings technology, as well as upcoming chip lines like the Blackwell Ultra and Rubin, with Feynman processors slated for 2028.
These advancements reflect Nvidia’s strategic focus on both building and running AI models, addressing the evolving needs of the tech industry.
ALSO READ | OpenAI’s Massive Abu Dhabi Data Center: A Game-Changer Bigger Than Monaco
Computex 2025: A Global Stage Amid Trade Tensions
Computex 2025 serves as the first major gathering of tech leaders since U.S. President Donald Trump threatened sweeping tariffs to boost domestic production. The event, attended by thousands, highlights the resilience of Asia’s tech supply chain while showcasing innovations like NVLink Fusion.
Huang’s presence continues to draw significant attention, with his 2024 Computex appearance dubbed “Jensanity” due to the fervent public response. This year, his keynote reinforces Nvidia’s role in shaping the future of AI and computing.
Comments (0)
Please sign in to leave a comment