Key Morningstar Metrics for Nvidia
- : USD 260Fair Value Estimate
- : ★★★★Morningstar Rating
- : WideMorningstar Economic Moat Rating
- : Very HighMorningstar Uncertainty Rating
Nvidia NVDA hosted its annual GTC conference and gave an outlook of USD 1 trillion of cumulative revenue from Blackwell and Rubin AI products from 2025-27. They also announced low-latency processors from their recent license with Groq, and NemoClaw to support OpenClaw “agentic” artificial intelligence.
Why it matters: There is simply no slowdown in AI spending or development on the horizon. The USD 1 trillion forecast implies both a stellar fiscal 2027 (effectively calendar 2026) and much more than USD 500 billion of data center revenue in fiscal 2028, ahead of FactSet consensus estimates.
- Nvidia’s forecast does not include contributions from “other” AI products, such as the Groq lineup for low latency, or storage or CPU racks. Nvidia estimates that Groq may make up 25% of inference workloads over time, so this excluded bucket might not be immaterial. The forecast also excludes Hopper GPU revenue (mostly in 2025) and Rubin Ultra GPU revenue (likely arriving in the second half of 2027).
- At the event, we were most impressed with the buzz around OpenClaw and Nvidia’s support behind the nascent, open-source agentic AI product. Agentic AI appears to be well on its way.
The bottom line: We raise our fair value estimate for wide-moat Nvidia to USD 260 from USD 240, based on an improved near-term forecast, even as we taper our long-term growth estimates.
- Shares appear materially undervalued, and we’re surprised by the market’s flattish stock price reaction. We believe this suggests that investors are skeptical of Nvidia’s forecast, although we have few concerns, barring an economic catastrophe such as further wars or financial crises.
- The AI ecosystem is still computing-constrained, and Nvidia’s forecast suggests that few, if any, of the leading cloud vendors will cut back on AI spending in the next two years. Given the development of agentic and physical AI, we foresee a long runway for AI gear spending.
