Key Takeaways

  • Crusoe is in discussions to raise about $3 billion at a valuation of roughly $30 billion.
  • The company's shift from flared-gas crypto mining to fully owned AI data center campuses is reshaping its market position.
  • Competition among CoreWeave, Lambda, Nebius, and Crusoe is intensifying as demand for GPU-rich infrastructure accelerates.

Crusoe's rapid valuation climb is drawing new attention across the AI infrastructure landscape. According to Bloomberg, the Denver-based company is in talks to secure around $3 billion at a valuation close to $30 billion. That figure would nearly triple the $10 billion valuation Crusoe reached in October 2025, when it closed a $1.38 billion Series E co-led by Valor Equity Partners and Mubadala Capital (source). The momentum is striking because the broader hardware and data center segment has become a hyper-competitive race to secure physical resources.

The founders, the chief executive and chief operating officer, launched the company in 2018 with an unusual operating model that turned flared natural gas from remote oil sites into power for Bitcoin mining. This early focus on energy management established foundational capabilities for its current operations.

Now the company is fully focused on building AI-native campuses. These facilities combine high-performance computing with purpose-built power generation, cooling systems, and cloud operations. It is a vertically integrated approach that avoids renting capacity inside someone else's data center, a choice that stands out in a sector where GPU inventory has become one of the most constrained inputs in the industry. Compute availability directly shapes the ability to train frontier models.

Forecasts from multiple firms point to an environment where AI workloads reshape global cloud economics. Projections from Gartner suggest public cloud services could reach $1 trillion in end-user spending by 2027, up from $563 billion in 2023. It is not just about more cloud consumption but about a deeper mix of specialized infrastructure tied to accelerated computing.

On the hardware side, IDC expects AI-centric infrastructure spending to grow at a 27% CAGR from 2023 to 2027. That includes accelerated servers, network fabrics tuned for large model training, and the data center systems that sit around them. This rapid expansion heavily favors operators with immediate access to high-density facilities.

Power availability remains a primary constraint in AI development. Analysts at McKinsey estimate that generative AI may drive 10% to 15% annual increases in data center power demand in leading markets. Crusoe addresses this power constraint by integrating energy generation directly with its data center infrastructure.

Crusoe reportedly has contracts to supply AI compute capacity to major hyperscale customers such as Oracle and Meta. The broader industry is seeing massive capital mobilization, from individual hyperscaler deployments to proposed initiatives involving OpenAI and SoftBank that envision hundreds of billions in AI data center investments. Hyperscalers and model developers are increasingly looking for partners who can natively control land, grid interconnection, construction, and operations.

Competitors are taking different approaches. Players like CoreWeave and Lambda similarly focus on GPU-rich, AI-optimized cloud and colocation capacity, often closely aligned with hardware providers like Nvidia. Other infrastructure firms like Nebius are also aggressively expanding their footprint to capture surging demand. These differences in posture, ownership, and financing create a varied competitive map to determine which model will scale more predictably as demand accelerates.

Standards are shaping how these operators build and run infrastructure. Kubernetes from the Cloud Native Computing Foundation has become the default orchestration engine for AI workloads that need portability across clusters and regions. Operators also track metrics from guidelines such as ISO/IEC 30134 to measure energy use, cooling efficiency, and other data center KPIs. These frameworks establish baseline metrics for comparing facility efficiency and maintaining operational consistency.

Investors appear to be reading these signals and making bets accordingly. A company that owns its power sourcing, construction timeline, and compute layer can sometimes mitigate risks that arise when supply chains tighten. The reported $30 billion valuation reflects a broader industry trend where control over physical AI infrastructure is shaping up to be just as strategically important as the model innovation happening on top of it.

The companies that secure power, land, and GPUs at scale aim to capture the highest-margin contracts from top-tier model developers. Crusoe is actively expanding its full-stack model to capture this market share.