NVIDIA leads the AI revolution: from robotaxis to 6G networks, GTC 2025 illuminates the tech frontier
On October 28, 2025, NVIDIA's GTC (GPU Technology Conference) 2025 in Washington, D.C., reached its climax. The annual gathering brought together global AI leaders, developers and corporate executives to focus on "the future of AI infrastructure." In his keynote, NVIDIA CEO Jensen Huang stressed that AI is moving from "laboratory innovation" to "global deployment," and that the company is consolidating its dominance in data centers, autonomous driving and networking through a series of strategic partnerships and product launches. Highlights included a robotaxi initiative with Uber, a 6G AI‑RAN push with Nokia, and the debut of the DGX Spark mini supercomputer. These moves not only pushed NVIDIA's stock up 5% intraday but also foreshadow an expected $150 billion increase in data‑center revenue by 2026, signaling AI's shift from concept toward a trillion‑dollar industry.
Uber Robotaxi: NVIDIA powering a global autonomous mobility ecosystem
One of the conference's most attention‑grabbing announcements was a deep collaboration between NVIDIA and Uber to build the world's largest robotaxi network. NVIDIA said it will provide Uber with the DRIVE AGX Hyperion platform to support deployment of 100,000 autonomous vehicles, with a planned rollout in 2027. In a live demo, Jensen Huang showed the platform integrating Blackwell GPUs and Grace CPUs to process multi‑sensor data in real time and achieve L4 (highly automated) driving capability. Uber CEO Dara Khosrowshahi said the partnership will "reshape urban mobility," noting Uber's global network will for the first time integrate human drivers and robot drivers across North America, Europe and Asia.
The project is part of a broader expansion of NVIDIA's autonomous driving ecosystem. Tom's Guide reported that NVIDIA's DRIVE platform has already extended to automakers including Stellantis, Lucid and Mercedes‑Benz. Stellantis plans to build autonomous vehicles on its LCV and STLA Small AV‑Ready platforms integrating NVIDIA technology; Lucid Motors is focusing on AI optimization for premium EVs; and Mercedes‑Benz emphasizes safety redundancy to handle complex driving scenarios. Foxconn will serve as a manufacturing partner responsible for vehicle assembly to help localize the supply chain. The alliance is expected to cover major global cities.
NVIDIA's autonomous ambitions stem from its computing platform advantage. DRIVE AGX Hyperion supports end‑to‑end AI training, from sensor fusion to path planning, providing a full‑stack autonomous solution. Huang said the project will "make every vehicle a mobile data center," and that the global robotaxi market could reach $2 trillion by 2030. Challenges remain, however: regulatory approvals (such as U.S. NHTSA standards), safety incident risks and ethical issues like employment impacts will test deployment timelines. Uber said it will pilot services in Los Angeles and San Francisco starting in 2027, with gradual expansion thereafter.
Nokia 6G AI‑RAN partnership: reshaping AI network infrastructure
Another major announcement was a strategic alliance between NVIDIA and Nokia. NVIDIA said it will invest $1 billion to acquire a 2.9% stake in Nokia to advance development of 6G AI‑RAN (Radio Access Network). The investment marks NVIDIA's deeper foray from chips into telecom infrastructure; Nokia's stock jumped 15% after European trading opened and at one point rose over 20.33%, hitting a two‑year high. Huang explained that the AI era requires "intelligent networks," and that Nokia's optical modules and 5G/6G technology will integrate with NVIDIA's Spectrum‑X platform to enable low‑latency, high‑bandwidth AI transport.
Spectrum‑X is NVIDIA's Ethernet AI networking platform optimized for data centers, supporting terabyte‑level bandwidth and nanosecond‑level latency. With Nokia, the platform will extend into telecommunications to support AI‑RAN—an AI‑driven radio access network that can dynamically allocate resources and reduce power consumption by about 30%. Nokia CEO Pekka Lundmark said the alliance will "accelerate the transition from 5G to 6G," with a target for commercial 6G around 2028 to support edge AI use cases such as real‑time video analysis. Analysts predict the collaboration could generate an additional $50 billion in orders for NVIDIA and drive a $150 billion increase in data‑center revenue by 2026, accounting for roughly 70% of total revenue.
The investment also reflects geopolitical strategy: partnering with Nokia, a European giant, helps NVIDIA mitigate exposure to U.S.–China chip tensions. However, EU antitrust review could delay completion, which is expected in Q4.
DGX Spark mini supercomputer: a "pocket jewel" for local AI
The conference also reiterated the DGX Spark launch — a mini supercomputer first announced two weeks earlier — designed for high‑speed local AI model execution. The desktop‑sized device integrates the Blackwell architecture, with peak token generation rates more than five times those of traditional GPUs, and supports inference for large models such as Llama‑70B. Huang demonstrated that DGX Spark can run complex AI tasks on‑premises, such as real‑time data analysis or edge computing, at roughly one‑third the power consumption of conventional servers. Tom's Guide said the product targets small and medium enterprises, with a price around $3,999 (excluding tax), and could accelerate a "de‑clouding" trend.
DGX Spark's innovation lies in its NIM microservice framework, which enables seamless deployment of containerized AI models. Combined with the Uber project it could be used for in‑vehicle AI training; together with the Nokia partnership it could support distributed 6G compute. Shipments are expected to exceed 100,000 units in 2026, contributing about $20 billion in revenue.
Conference impact: global radiation of the AI ecosystem
GTC 2025 attracted 25,000 onsite attendees and millions online, highlighting NVIDIA's industry pull. NVIDIA's market value rose by more than $90 billion intraday. These partnerships suggest AI is penetrating from cloud to edge, with the global AI hardware market projected to reach $500 billion by 2026. Environmental challenges are prominent: data centers already account for about 2% of global electricity consumption and may rise further. (NVIDIA has pledged to follow a "green AI" framework.)
Note: On October 29, 2025, NVIDIA became the world's first company with a market capitalization exceeding $5 trillion; the stock rose 3% to $207 (intraday high $212.19), and market value jumped from $4.89 trillion to $5.03 trillion. This milestone was driven by AI chip demand, far outpacing Apple and Microsoft, though bubble risks remain.
Related Articles:
NVIDIA Investments and Collaborations (2023–2025): 127+ Cases and Key Moves (as of 2025‑10‑29)
NVIDIA Invests $1 Billion in Nokia, Triggering Share Surge
AI Bubble Doubts: Tech Insiders Call It "Overhyped" as Capital Frenzy Continues
Qualcomm Makes Bold Push into AI Chips; Two 2026 Models Target NVIDIA's Data‑Center Dominance