December 23, 2025 • 2 min read DEEP TECH

NVIDIA: The Industrial Revolution of Intelligence

Demand for Blackwell is not just 'strong'—it is structural. We are moving from general purpose computing to accelerated computing, and NVIDIA has a 10-year lead.

We are in the early innings of a platform shift that rivals the internet itself. Reviewing NVIDIA’s position going into 2026, the data points to a supply-constrained environment that will last for years.

The Supply/Demand Dislocation

Jensen Huang describes the demand for Blackwell as “insane,” but the numbers tell a clearer story.

  • Revenue Visibility: Reports suggest $500 Billion in line-of-sight revenue through 2026.
  • Data Center Dominance: $51.2B in segment revenue (+66% YoY) confirms that the Data Center is effectively the new “unit of compute.”
  • Analyst Consensus: The “Strong Buy” rating remains almost unanimous (39 of 41 analysts), with price targets continually revised upward to refect the durability of this capex cycle.

Agentic AI: The Next Multiplier

The market is pricing NVIDIA on training demand. It is largely ignoring inference demand.

As we move from “Chatbots” to “Agentic AI” (systems that reason and take action), inference compute costs will ballon by 10-100x.

  • Blackwell Architecture: Designed specifically to handle this increasing inference load.
  • The Moat: It’s not just the chip; it’s CUDA, NVLink, and the NIM microservices that make deployment possible.

Sustainability of Growth

Bear cases focus on “Over-ordering” or “Cyclicality.”

MD&A Rebuttal: Management highlights that Hyperscalers (Microsoft, Meta, Google) are essentially in an arms race. To stop buying GPUs is to capitulate on the future of their core products. This is not inventory building; it is infrastructure building.

Conclusion

NVIDIA is building the railroads of the 21st century. Until we see a slowdown in the deployment of intelligence, we remain overweight.

Related Thoughts

Dec 23

ABB: Electrifying the Physical World

← Back to Thoughts