Leopold Aschenbrenner's Situational Awareness Fund Just Bought This Nvidia-Backed Neocloud Stock

Nebius is benefiting from surging AI infrastructure demand, with first-quarter revenue up 684% year over year and EBITDA and adjusted EPS beating expectations by 43% and 50%, respectively. The company also secured a $2 billion Nvidia investment commitment and a $2.6 billion Bloom Energy power deal, reinforcing its long-term AI build-out. The article is constructive on Nebius’s growth and profitability trajectory, though it flags execution, competition, and power-supply risks.

Analysis

This is less a pure NBIS rerating story than a signal that the AI infrastructure capex cycle is becoming self-reinforcing. The strategic implication is that compute supply is no longer just constrained by GPUs; it is increasingly gated by power, networking, and deployment execution, which should extend pricing power for the best-capitalized neoclouds while squeezing smaller entrants that cannot finance multi-year buildouts. That dynamic favors suppliers with balance-sheet leverage to the theme, but it also raises the odds of future bottlenecks migrating from chips to energy interconnects and data-center commissioning.

The market is underestimating how much Nvidia’s involvement de-risks NBIS’s roadmap and, more importantly, how it can accelerate customer adoption through ecosystem validation. If Nebius can keep turning capex into contracted utilization, the stock can remain optically expensive on near-term P/S while still being cheap on forward capacity growth; the real bull case is not revenue alone, but improved return on incremental invested capital as each new tranche of capacity is pre-sold. The main second-order beneficiary is Bloom Energy, because AI power demand increasingly shifts from a utility-cost line item to a strategic input where on-site generation and grid optionality become critical.

The contrarian risk is that consensus is extrapolating scarcity economics too far out on the curve. Once hyperscalers and large-cap neo-infrastructure peers bring more capacity online over the next 12-24 months, GPU rent and inference margins can compress faster than top-line growth slows, especially if utilization normalizes from shortage levels to merely strong levels. In that scenario, the market could de-rate the whole cohort even if revenue keeps compounding, because investors will start valuing these names on mid-cycle margins rather than peak-scarcity multiples.

Key Decisions for Investors

• Long NBIS on 3-6 month pullbacks, but size it as a momentum/optionality position rather than a core fundamental holding; the setup works if utilization stays tight through the next two earnings prints, but the left-tail is multiple compression if capacity comes online faster than demand.
• Pair trade: long NBIS / short a basket of hyperscaler capex proxies with slower monetization latency over the next 6-12 months; the edge is that NBIS is closer to the scarcity rent while larger platforms may see capex drag before revenue catch-up.
• Add BE as a second-order beneficiary on any weakness over the next 1-3 months; the trade is not earnings beta, it is the growing strategic premium on power-secured AI buildouts if grid constraints persist.
• For risk-managed exposure, buy NVDA call spreads 6-12 months out; Nvidia benefits even if NBIS-specific economics compress because the ecosystem validation and hardware cycle remain intact, but upside should be capped to reduce paid premium.
• Avoid chasing NBIS after large gap-ups; wait for a 10-15% retracement or a post-earnings setup, because this name is likely to trade on sentiment and capacity headlines more than fundamentals until the market sees evidence of durable margin stability.