Back to News
Market Impact: 0.3

CoreWeave vs. Nebius: Which AI Infrastructure Stock Is the Better Buy?

Source: Nasdaq

Artificial IntelligenceTechnology & InnovationCompany FundamentalsCorporate EarningsAnalyst InsightsCapital Returns (Dividends / Buybacks)
CoreWeave vs. Nebius: Which AI Infrastructure Stock Is the Better Buy?

AI cloud infrastructure spend is projected to top $200B by 2028, benefiting CoreWeave (CRWV) and Nebius (NBIS), though with different risk profiles. CRWV’s latest quarter revenue rose 420% YoY to $981.6M (+15.2% vs. consensus), but investors face heavy capex ($20B–$23B in 2025) and elevated interest expense ($264M in Q1; guided $260–$300M), alongside customer concentration (77% of 2024 revenue from top two customers). NBIS delivered 385% YoY revenue growth in Q1 and estimates ~5% higher commercial node availability after Slurm upgrades, but remains unprofitable with adjusted EBITDA negative for full-year 2025. In the past month, CRWV surged 189.6% and NBIS rose 66.9%, with Zacks framing NBIS as the more compelling pick despite near-term margin drag.

Analysis

The market is correctly assigning value to AI demand, but it may be misassigning who captures the economics. CRWV and NBIS are growth assets with very different balance-sheet profiles: the scarce commodity is not revenue growth, it is funded capacity that can be monetized before financing costs compound. That makes CRWV the more fragile equity despite the stronger top-line momentum, because capex and interest are effectively a call option on utilization staying very high for multiple quarters.

NBIS looks cleaner on operating quality, but the real second-order winner across the stack is NVDA and adjacent infrastructure vendors, not the cloud wrappers. These providers are still dependent on NVIDIA silicon, networking, and power density improvements, while public shareholders absorb execution risk, customer churn risk, and dilution risk. If specialized AI clouds continue to take share from AWS/Azure on inference and fine-tuning, the margin transfer is more likely to come from hyperscaler service mix than from core IaaS share.

The contrarian read is that the market may be overrating "best growth" as synonymous with "best stock." CRWV’s concentration and financing burden make it vulnerable to any slowdown in bookings or a dip in GPU utilization; NBIS is less levered operationally, but still needs cash conversion to prove its valuation. The near-term catalyst path is the next 1-2 earnings prints: what matters is gross margin, capex cadence, and whether interest expense starts to outrun incremental revenue. If free-cash-flow conversion does not improve by mid-2025, both names are likely to see multiple compression even if revenue keeps compounding.

AllMind Terminal

AI-powered research, real-time alerts, and portfolio analytics for institutional investors.

Request Trial

Market Sentiment

Overall Sentiment

mixed

Sentiment Score

0.05

Ticker Sentiment

CRWV0.15
NBIS0.25
NVDA0.10

Key Decisions for Investors

  • Prefer NVDA over CRWV as the cleaner AI infra expression for the next 3-6 months: long NVDA / short CRWV on rallies, targeting a lower-volatility way to participate in AI capex while avoiding customer-concentration and financing risk.
  • Do not chase CRWV after the recent vertical move; use any post-print strength to hedge or trim. Falsifier: if CRWV shows two consecutive quarters of improving gross margin and lower interest burden versus guidance, the short becomes less attractive.
  • Watch-list long NBIS only on evidence of cash-burn improvement and customer diversification; otherwise treat it as a 'prove-it' name. A better entry is after the next earnings update if adjusted EBITDA guidance is reaffirmed and capex remains funded without dilution.
  • Long the power/cooling/data-center beneficiaries rather than the most levered cloud wrappers: pair AI infra exposure with VRT or EQIX over a 6-12 month horizon if you want exposure to the buildout without taking direct financing risk.
  • Set an alert for any CRWV/NBIS guidance cut or debt-market widening; a 100-150 bps move higher in financing spreads or a slowdown in backlog conversion would likely compress multiples faster than headline revenue can support them.

More News

From AllMind Research

Browse all research