Nvidia Invested $2 Billion in This Artificial Intelligence (AI) Stock. It Is Already Up 57% and Can Become a Multibagger

Nebius has a $46 billion revenue backlog, is expanding capacity with a new 310 MW AI factory in Finland, and targets 3 GW of contracted power by end-2026 versus 170 MW of active capacity at the end of 2025. Management’s partnership with Nvidia to build physical-AI-focused data centers and software support underpins the case for multibagger upside, with revenue rising from $530 million last year and possible 2028 revenue of $16 billion. The article is broadly bullish on NBIS, though it is framed as a stock-picking opinion piece rather than new company-reported results.

Analysis

Nebius is becoming less of a standalone growth story and more of a financing-and-execution call on the AI infrastructure buildout. The important second-order effect is that Nvidia’s investment and technical alignment likely compress Nebius’ cost of capital while also making it a preferred channel for customers who want a de-risked path to inference capacity; that can accelerate bookings conversion faster than a typical neocloud peer. The market is probably still underestimating how much of this story is about scarcity of power and delivered capacity, not just headline demand.

The near-term winner is clearly NVDA: every incremental partner datacenter expands Nvidia’s platform lock-in and creates a software-plus-hardware wedge that is harder for hyperscalers to disintermediate. META and MSFT are indirect beneficiaries if Nebius becomes a capacity overflow valve, but they also gain bargaining leverage versus incumbent cloud providers by showing that alternative inference supply is scaling. The competitive loser is the broader long-duration, non-Nvidia accelerated-compute stack: if inference workloads get standardized around Nvidia-anchored, pre-integrated environments, smaller AI cloud providers without power access or strategic chip relationships will face a funding and utilization squeeze.

The main risk is that the market is extrapolating contracted power into monetizable revenue too quickly. Power procurement, grid interconnection, and build-out timelines can slip 6-18 months, and any delay would force Nebius to bridge capex with dilution or debt before backlog converts. A second-order risk is customer concentration: if a few large hyperscalers reprice or internalize more inference demand, backlog quality can deteriorate even while reported bookings stay elevated.

Contrarian angle: the move may be overowned in the short run because the stock is trading like capacity is already fungible and fully monetizable, when in reality the bottleneck is deployment execution. The better asymmetric setup may be NVDA versus NBIS—NVDA captures option value across every successful partner, while Nebius bears the capital intensity and project risk. If the market starts rewarding actual delivered MW over announced contracted MW, NBIS could still work, but volatility should stay high until the next 2-3 quarters of capacity milestones are hit.

Key Decisions for Investors

• Long NVDA / short NBIS as a 3-6 month pair trade: NVDA retains upside from ecosystem expansion with lower execution risk, while NBIS is more exposed to build-out slippage and capex financing; target a 1.5-2.0x gross exposure with a hard stop if NBIS successfully de-risks major power milestones.
• Buy NBIS on post-rally pullbacks only, not strength: use a 30-60 day entry window and look for entries after any 10-15% retracement tied to financing or construction headlines; the risk/reward improves if the market overreacts to timing noise rather than backlog quality.
• Accumulate NVDA on any weakness tied to neocloud skepticism: this is a lower-volatility way to express the same AI infrastructure thesis, with better downside protection because Nvidia monetizes the ecosystem regardless of which operator wins share.
• Optionality trade on NBIS: buy 6-12 month call spreads rather than stock if you want exposure to capacity ramp upside but want to cap downside from project delays; this fits a catalyst path where milestone updates matter more than quarterly earnings.
• Watch META and MSFT as secondary beneficiaries: if alternative inference capacity scales, their cloud bargaining power improves; use any relative underperformance in META/MSFT versus NBIS/NVDA as a signal that the market is over-discounting the strategic supply option.