From Shanghai to Taipei, Jensen Huang is reshaping Nvidia’s global AI strategy — one chip, one hub, one continent at a time. At Computex 2025, Nvidia CEO Jensen Huang delivered a blockbuster keynote packed with new announcements, global ambitions, and one clear message: Asia is central to the AI future.
New Tech Unveiled at Computex 2025
NVLink Fusion
- Nvidia launched a new version of its chip-to-chip interconnect tech — NVLink Fusion — that allows chips from other companies to connect with Nvidia’s systems.
- It’s a bold shift: Nvidia is now opening its ecosystem to rival chipmakers like Marvell and MediaTek, enabling more custom AI systems.
DGX Spark Desktop AI PC
- Huang confirmed Nvidia’s AI PC for researchers is now in full production and will ship in a few weeks.
- Built for desktop AI development — from training to inference — the Spark system is being co-developed with Acer, Asus, and Gigabyte.
Nvidia Constellation HQ – Taipei
- Huang announced plans for a new headquarters in Beitou-Shilin, Taipei, calling it “one of the largest products we’ve ever built.”
- The new hub will anchor Nvidia’s expanding R&D and AI hardware operations in Taiwan.
Next-Gen Chips Roadmap
- Huang teased the roadmap beyond the Blackwell Ultra (coming later this year):
• Rubin chips are on deck next.
• Feynman processors will follow — arriving around 2028. - These chips aim to support the shift from training AI to running real-world AI applications.
Global Strategy: Made in Asia, Powered by Nvidia
China: Nvidia is opening a new R&D center in Shanghai to stay close to customers despite U.S. export restrictions. Chinese officials are offering tax breaks and regulatory support, while Nvidia works on downgraded versions of H20 chips and a new Blackwell-based chip to stay compliant.
Taiwan: Taiwan becomes Nvidia’s AI production hub:
• Partnering with Foxconn and TSMC to build a national AI supercomputer.
• Supporting Taiwan’s AI and robotics ecosystem with infrastructure and systems.
What’s the Play?
Opening NVLink Fusion to competitors is bold — it positions Nvidia as the platform leader, not just a chip designer. R&D expansion in China keeps Nvidia close to a critical (but restricted) market without violating U.S. rules. Doubling down in Taiwan anchors manufacturing and innovation near key supply chain allies like TSMC and Foxconn.
Jensen Huang is playing a long game. While Washington imposes export controls, Nvidia is quietly building the AI future in Asia.
- China gets a tailored, compliant Nvidia chip strategy.
- Taiwan becomes the AI hardware factory of the world.
- And NVLink Fusion opens Nvidia’s ecosystem to others — solidifying its central role in global AI infrastructure.
This isn’t just a product launch. It’s a global repositioning.
Disclosure: This article does not represent investment advice. The content and materials featured on this page are for educational purposes only.
Related:
Earnings Calendar, Fed Speech, Housing Data: What to Watch This Week
UK overtakes China as second-largest US Treasury holder
CERN Scientists Turn Lead into Gold — Why This Is Bigger Deal for Bitcoin Than You Think
Nvidia, Cisco, Oracle and OpenAI are backing the UAE Stargate data center project
Trump vs. Tim Cook: Apple’s Global Strategy Sparks Debate Over Economic Patriotism and AI Security
US and EU break impasse to enable tariff talks
China Tightens Control Over AI Data Centers
Trump: US will set new tariff rates, bypassing trade negotiations
Sundar Pichai Interview on AI, Search, and Future of Google
Billionaire Investors Reveal Q1 2025 Portfolio Moves: Buffett, Ackman, Tepper, Burry & More
