Unprompted is an occasional opinion column from Kunal Gupta for Pivot 5 readers.

Unprompted: Beyond Chips and Clouds

This week, I found myself in a sea of 25,000 attendees at NVIDIA's GTC conference in San Jose, California. 

People were excitedly discussing everything from large language models to autonomous vehicles. But it was during Jensen's keynote that something clicked for me. As Jensen casually called a robot on stage and mentioned, almost as an afterthought, that it contained two GPUs, I realized we've been thinking about NVIDIA all wrong.

When we talk about NVIDIA, the conversation invariably gravitates toward data centers, cloud computing, and those perpetually sold-out chips. Their astronomical stock performance reinforces this narrative. But what struck me at GTC wasn't the cloud story—it was seeing NVIDIA quietly extending its reach into places I haven’t heard many talk about.

Beyond the Cloud

Standing in a demo hall, I watched a grocery store system that uses computer vision to recognize customers and instantly customize advertisements on in-store displays. Inside those unassuming fixtures? NVIDIA chips. 

Then there was NVIDIA’s new AI computer for engineers and data scientists, promising to bring GPU computing directly to knowledge workers' desks. 

Despite the vast majority of NVIDIA's revenue flowing from data centers—with cloud providers such as Google, Microsoft, Amazon and Meta, as the biggest customers—what's not making headlines is how their chips are steadily infiltrating our physical world. From manufacturing floors to retail environments to autonomous machines, NVIDIA chips are becoming the foundation for AI at the edge.

The question isn't just about how many NVIDIA can ship anymore. The more fascinating question is: are we witnessing the birth of a computing paradigm where specialized AI processing becomes as ubiquitous and essential as electricity?

Beyond the Chip: The Full-Stack Play

What truly separates NVIDIA from would-be competitors isn't just superior silicon—it's their increasingly comprehensive ecosystem. While analysts obsess over thel chips, NVIDIA has quietly assembled something potentially more valuable: a complete technology stack.

Their CUDA platform has become the de facto standard for AI development, creating powerful network effects that even the most determined competitors struggle to overcome. But what fascinated me most at GTC was Omniverse, their platform for creating AI-powered simulations of physical environments.

I took a virtual tour of a BMW factory—not a physical one, but a digital twin recreated in Omniverse with stunning fidelity. BMW engineers use this AI-simulated environment to make real-time decisions about maintenance, optimize throughput, and enhance process efficiencies. The virtual factory responds and evolves like its physical counterpart, creating a feedback loop that continuously improves both.

This isn't just impressive technology; it's a strategic masterclass. While Intel focuses on chips and Microsoft on software, NVIDIA has built an integrated solution connecting hardware, development tools, and application environments. The parallels to Apple's strategy are unmistakable—create a tightly integrated ecosystem where hardware, operating system, and application layer work in concert, making it simultaneously easy to develop for and difficult to leave.

The Invisible Moat

What makes NVIDIA's approach particularly brilliant is how each layer reinforces the others. Their GPUs power the CUDA platform, which makes developing AI applications more accessible, driving more GPU adoption. Omniverse creates demand for specialized simulation capabilities, which—surprise—NVIDIA's hardware handles exceptionally well.

This cycle has created an invisible moat around NVIDIA that competitors will find increasingly difficult to cross. AMD and Intel may build competitive chips, but they lack the software ecosystem. Cloud providers may create custom silicon, but they don't have NVIDIA's developer mindshare.

The dynamics are similar to Microsoft's Windows dominance in the 90s or Apple's mobile ecosystem in the 2010s. NVIDIA isn't just selling components—they're becoming infrastructure for the AI age.

The Strategic Implications

For business leaders thinking about their AI strategy, NVIDIA's evolution suggests several key insights:

1. Full-stack thinking wins: The companies that integrate hardware, software, and platforms will create more value than those focusing on just one layer.
   

2. AI at the edge is accelerating: The proliferation of NVIDIA chips beyond data centers signals that AI processing is following computing's historical pattern of moving closer to the point of use.
   

3. Developer ecosystems matter enormously: NVIDIA's CUDA advantage demonstrates how developer adoption creates compounding competitive advantages.
   

4. Simulation becomes reality: Digital twins like BMW's factory in Omniverse show how the lines between physical and digital operations are blurring, creating new opportunities for optimization.
   

The Road Ahead

The NVIDIA I saw at GTC isn't just a chip company riding the AI wave—it's transforming into a full-stack AI infrastructure provider. While everyone focuses on how many GPUs they can manufacture, the more interesting story is how they're weaving themselves into the fabric of AI development and deployment.

As Jensen closed his keynote, surrounded by the latest hardware but talking mostly about software and applications, it became clear that NVIDIA's ambitions extend far beyond being the "picks and shovels" provider for the AI gold rush. They're building the mines, the transportation networks, and even helping design what gets extracted.

The question for the rest of the industry isn't just how to compete with NVIDIA's chips—it's how to respond to a company that's systematically eliminating the boundaries between hardware, software, and applications in the AI stack. 

For those of us watching this unfold, it's a masterclass in strategic positioning that will shape how AI gets built and deployed for years to come.

Unprompted is an occasional opinion column from Kunal Gupta for Pivot 5 readers. Follow Kunal on LinkedIn.