NVIDIA CEO Jensen Huang launched COMPUTEX 2025 with a powerful keynote address packed with significant announcements, groundbreaking technologies, and innovative partnerships designed to shape the future of AI, computing, and robotics. Here’s a detailed overview of the key highlights from the event.
NVLink Fusion: Next-Level AI Infrastructure
Jensen Huang introduced NVIDIA NVLink™ Fusion, a revolutionary silicon technology that allows hyperscalers to create semi-custom, ASIC-hybrid AI infrastructures. NVLink Fusion significantly enhances scalability and performance, enabling unprecedented AI computing power. Major industry players and NVIDIA’s ecosystem partners are poised to leverage this innovation for expansive AI deployments.
NVIDIA RTX PRO Servers: Accelerating AI Integration in Enterprises
NVIDIA unveiled RTX PRO™ Servers alongside the NVIDIA Enterprise AI Factory validated design. These servers represent a critical advancement in enterprise computing, accelerating the adoption of AI-driven processes across design, engineering, and business applications. Enterprises globally will benefit from boosted productivity and transformative AI capabilities.
Foxconn and NVIDIA Partnership: Building Taiwan’s AI Future
In a landmark collaboration, Foxconn joined forces with NVIDIA and the Taiwanese government to develop a cutting-edge AI factory. Leveraging NVIDIA’s advanced Blackwell infrastructure, this facility will empower startups, researchers, and industries to innovate rapidly, driving Taiwan’s competitive advantage in global AI leadership.
Launch of ABCI-Q: The Quantum Research Supercomputer
Partnering with Japan’s National Institute of Advanced Industrial Science and Technology (AIST), NVIDIA launched ABCI-Q, the world’s largest quantum research supercomputer. With over 2,000 NVIDIA H100 Tensor Core GPUs integrated alongside quantum processors, ABCI-Q promises to revolutionize industries including healthcare, finance, energy, and beyond.
DGX Spark and DGX Station Systems: AI-First Personal Computing
NVIDIA announced global collaborations with Taiwan’s leading system manufacturers to launch the NVIDIA DGX™ Spark and DGX Station™ systems. These innovative personal computing systems are optimized for AI tasks, enabling data scientists, researchers, and developers to achieve superior efficiency and unprecedented computational performance.
Cloud-to-Robot Platforms: Advancing Humanoid Robotics
The keynote also highlighted significant strides in robotics, with NVIDIA unveiling Isaac™ GR00T N1.5, the enhanced humanoid reasoning foundation model, and Isaac GR00T-Dreams for generating synthetic motion data. These technologies, combined with NVIDIA Blackwell systems, promise to dramatically speed up humanoid robot development, setting new standards for robotic capabilities.
Grace CPU C1: Expanding AI Support Across Industries
The NVIDIA Grace™ CPU C1, designed for high-performance computing at the edge, telco, and storage sectors, received robust endorsement from numerous Original Design Manufacturer (ODM) partners. This highlights a significant market interest and adoption trajectory for Grace CPUs, essential for edge computing infrastructures.
ASUS-NVIDIA Collaboration: A Leap in Quantum and AI Research
ASUS has constructed a new supercomputer powered by NVIDIA technologies for the National Center for High-Performance Computing in Taiwan. Designed to propel quantum research, climate science, and large language models, this supercomputer solidifies Taiwan’s role as a leader in AI-driven research.
COMPUTEX 2025: Shaping the Future of Technology
This year’s COMPUTEX demonstrated NVIDIA’s relentless drive toward AI integration, quantum computing breakthroughs, and advanced robotics. Jensen Huang’s keynote highlighted the strategic importance of collaborations and cutting-edge hardware advancements that continue to position NVIDIA as an influential force in the tech world.
Watch the NVIDIA Keynote Replay
To experience Jensen Huang’s COMPUTEX 2025 keynote, watch the replay.
