2 Artificial Intelligence (AI) Infrastructure Stocks Outgrowing Nvidia

Nebius reported Q1 revenue growth of 684% year over year, while CoreWeave grew revenue 112% to $2.1 billion and now has nearly a $100 billion backlog, with over one-third expected to convert to revenue in the next two years. Wall Street expects Nebius revenue to rise from $530 million in 2025 to $11.2 billion by 2027, and CoreWeave revenue to grow 147% in 2026 and 97% in 2027. The article argues both AI neocloud names are outgrowing Nvidia, though neither is profitable and both rely on capital raises for expansion.

Analysis

The market is increasingly treating neoclouds as capacity-constrained utility assets rather than software businesses, which is why the equity story is now being driven more by backlog conversion and financing capacity than by headline growth rates. That shifts the competitive axis toward whoever can secure power, land, and GPUs fastest; second-order winners include datacenter REITs, electrical equipment suppliers, and high-end networking vendors, while smaller AI infra providers without balance-sheet access risk being squeezed out or forced into dilutive capital raises.

The key risk is that these businesses are front-loading capex into a demand curve that may remain strong but becomes harder to translate into equity returns if cost of capital stays elevated. In practice, the first-order growth narrative can coexist with poor stock performance if customer concentration, GPU depreciation, or refinancing spreads widen over the next 6-18 months. The backlog headline is bullish, but backlog quality matters: if a meaningful share is tied to a few hyperscale or AI model customers, any scheduling slippage would hit revenue recognition and sentiment simultaneously.

Consensus is probably underestimating how much this wave benefits Nvidia indirectly even if the market is focused on NBIS and CRWV. The real option value sits with the upstream enablers: every incremental rack deployed by neoclouds pulls through more GPUs, networking, optics, and power infrastructure, so NVDA remains the cleaner expression of the demand shock with less financing risk. The contrarian take is that NBIS/CRWV are momentum vehicles on infrastructure buildout, but the better risk-adjusted trade may be to own the picks-and-shovels while fading the most diluted capital structures.

Near term, watch for equity issuance, debt pricing, and any revision in backlog conversion timing; those are the catalysts that will matter over days to months. Over years, if utilization stays high and financing remains open, both names can compound sharply, but the path will likely be volatile and highly sensitive to market liquidity.

Key Decisions for Investors

• Long NVDA vs. basket long NBIS/CRWV: favor the supplier of constrained inputs over the balance-sheet risk takers; best risk/reward if AI capex remains strong over the next 6-12 months.
• If expressing bullishness on the neocloud theme, size a small tactical long in CRWV over NBIS for the next 1-3 quarters: backlog visibility is better, but use a tight stop because equity dilution risk is higher.
• Avoid initiating fresh outright longs in NBIS ahead of any capital raise window; wait for secondary financing and buy post-dilution if enterprise value/revenue resets lower.
• Overweight AI infrastructure enablers (e.g., NVDA and related power/networking beneficiaries) into any pullback: they capture the same spend with materially lower financing risk and better margin durability.
• For a more defensive expression, pair long NVDA against short a basket of high-multiple, cash-burning AI infra names if credit spreads widen; this is a relative-value trade on cost of capital, not on AI demand itself.