Industry is going through what Nvidia CEO Jensen Huang describes as a major reset. In his latest keynote, he argued that the world is entering what may become the largest infrastructure build‑out in human history, driven by demand for AI and accelerated computing. Around the globe, companies are constructing AI factories, semiconductor fabrication plants and large‑scale computing facilities — and many of the organisations behind that build‑out were represented at Nvidia’s GTC conference.
Huang’s remarks situate Nvidia not just as a supplier of chips and software, but as a central player in a cross‑industry investment cycle that spans healthcare, media, quantum research, retail, robotics, manufacturing and telecommunications. Each of these sectors, he said, is reaching its own “deep learning and transformer moment”, where systems begin to find patterns and generate insights automatically from large datasets.
Healthcare, media and real-time AI platforms
In healthcare, Huang described a “ChatGPT moment” as AI starts to transform research, diagnostics and patient support. That momentum feeds back into infrastructure demand, as hospitals, research institutions and biotech firms look for computing platforms that can handle drug discovery simulations, clinical decision support agents and medical robotics.
Media, entertainment and gaming are also leaning heavily into AI. At GTC, Nvidia highlighted real‑time AI platforms used for live translation, broadcasting, game environments and video production. As more content is assisted or enhanced by AI — from generated scenes to adaptive game worlds — the need for powerful, low‑latency compute continues to climb.
Quantum-GPU hybrids as part of the next wave
Huang pointed to quantum computing as another front in this infrastructure boom. Nvidia is working on hybrid systems that combine GPUs with quantum processors, with more than thirty companies collaborating on next‑generation quantum‑GPU hybrid computing systems.
While large‑scale quantum computers are still an emerging technology, the decision to pair them with GPUs reflects a pragmatic view: many useful workloads will likely need classical and quantum resources working together. Nvidia’s role is to provide the accelerated classical side and the software layers that allow developers and researchers to orchestrate tasks across both parts of the system.
Retail, CPG, robotics and manufacturing
Retail and consumer packaged goods (CPG) represent roughly a $35 trillion global industry, according to the figures Huang cited. Nvidia’s technology is used there for supply chain optimisation, recommendation systems and AI shopping agents that assist customers and support service teams. In sectors that operate on thin margins and high volume, even small efficiency gains can translate into significant absolute impact.
Robotics and manufacturing, which Huang described as roughly a $50 trillion industry, have been a focus for Nvidia for more than a decade. The company concentrates on three fundamental computing systems for robotics: systems for training robots in simulation, systems for running AI models inside the robot, and systems for operating robotic fleets in real environments. At GTC, dozens of robots built by partners were demonstrated, underscoring how widely Nvidia’s platforms are being integrated into physical machines.
These efforts tie directly into the infrastructure story. Training simulators, onboard inference engines and fleet management services all require specialised compute, often deployed across data centres, edge devices and factory floors.
Telecoms and the shift to AI-powered base stations
In telecommunications, an industry Huang estimated at around $2 trillion, Nvidia is targeting the evolution of mobile network infrastructure. In the previous generation of computing, the mobile network base station’s job was to handle wireless communication. In the future, Huang said, base stations will also become edge AI computing platforms, running AI workloads close to users.
Nvidia’s platform for this area is called Aerial. It accelerates 5G and future network workloads and is used in collaborations with partners such as Nokia and T‑Mobile. By turning base stations into edge AI nodes, operators can run applications such as real‑time analytics, low‑latency media processing and network optimisation nearer to end users rather than relying solely on distant data centres.
Why Huang calls this a historic infrastructure build-out
Across all of these sectors, the common thread is that AI is moving from isolated projects into core operations. Healthcare organisations, retailers, manufacturers, telcos and content producers are all standing up new computing facilities or retrofitting existing ones to handle deep learning, transformers and real‑time inference. Semiconductor fabs and AI factories are being built to supply that demand.
Huang’s argument is that this layered transformation — from chips and networks up through applications in trillion‑dollar industries — will make the current build‑out one of the largest infrastructure cycles the technology sector has ever seen. Nvidia’s role is to supply the accelerated computing platforms and software that sit at the heart of those deployments.
Sources
- Keynote remarks by Nvidia CEO Jensen Huang on AI infrastructure, industry reset and sector‑specific opportunities at GTC
- Nvidia GTC session materials on real‑time AI in media and gaming, quantum‑GPU hybrids, retail and CPG, robotics, manufacturing and telecommunications
- Industry reports and public data on the approximate size of retail, robotics, manufacturing and telecom markets referenced in Huang’s talk