NVIDIA Closes GTC Washington With Quantum–GPU Links, US-Made Blackwell, 6G Push, Robotaxis and New AI Supercomputers
At its GTC conference in Washington, D.C., NVIDIA laid out a broad agenda for strengthening America’s AI leadership—spanning quantum–GPU integration, U.S. manufacturing milestones for Blackwell, telecom partnerships aimed at 6G, autonomous mobility at scale, and a slate of new AI supercomputers for scientific discovery.
Blackwell manufacturing in the U.S.
Jensen Huang highlighted progress toward U.S. production of Blackwell-generation AI chips, marking a symbolic on-shore step for advanced accelerators. Domestic manufacturing and system assembly are positioned as part of a wider strategy to harden supply chains and shorten time-to-deployment for sovereign AI infrastructure.NVQLink: bridging quantum processors and GPU supercomputers
NVIDIA introduced NVQLink, an architecture for tightly coupling quantum hardware with GPU supercomputers. The goal: real-time control, calibration and hybrid quantum-classical simulations under a unified software stack. If successful, it could reduce overhead in error correction workflows and enable larger-scale experiments as labs move from hundreds to far more qubits.Telecom ambitions and a 6G platform
The company also announced a strategic move into next-gen radio access networks with a partner program aimed at “AI-native” 5G-Advanced and a path to 6G. For carriers, the pitch centers on energy-efficient AI acceleration at the edge, software-defined upgrades, and a common platform unifying inference, scheduling and network optimization.Autonomy: Hyperion 10 and robotaxi scaling
On the road to Level 4, NVIDIA’s Hyperion 10 reference architecture—compute, sensors and DRIVE OS—was presented as a blueprint to help automakers and mobility platforms converge on safety-critical stacks. A highlighted collaboration targets a global network of robotaxis beginning in 2027, with a roadmap that scales fleets and standardizes updates across partners.Digital twins and AI factories for industry
Industrial programs showcased “AI factories” and high-fidelity digital twins for robotics and production lines. The emphasis: compressing design-to-deployment cycles, simulating edge cases in photoreal environments, and letting software-defined plants roll out changes without halting physical lines—an approach positioned to relieve structural labor shortages.Seven new AI supercomputers for U.S. science
NVIDIA also highlighted collaborations to build out AI supercomputing capacity for U.S. national labs, including multiple new systems designed for scientific discovery, security applications and open research. The systems combine Blackwell-class GPUs, high-performance networking and modern AI stacks to accelerate workloads from climate modeling to materials and fusion research.Bottom line
From quantum links to telecom, autonomy and national-lab compute, GTC Washington framed NVIDIA’s strategy as a multi-sector platform push: align software, silicon and ecosystems so breakthroughs reinforce each other—and keep the flywheel of model progress, infrastructure investment and real-world deployment spinning.Editorial Team — CoinBotLab