Odaily Planet Daily report: Analyst qinbafrank posted on X that the latest financial reports from CoreWeave (CRWV) and Nebius indicate that the AI cloud computing (CSP) industry is entering a phase of rapid expansion, with the focus of competition shifting from merely providing GPU rentals to building AI infrastructure platforms encompassing computing power, software, data, and operational capabilities.
Currently, demand for AI computing power still significantly exceeds short-term deliverable supply. At the same time, pricing power for AI computing is strengthening, but price increases are concentrated primarily on high-value resources. CoreWeave reported that prices for various GPU computing SKUs rose by approximately 25% in July; Nebius disclosed that prices for previous-generation GPUs increased by more than 30% compared to Q1, with average annualized revenue from new contracts in Q2 exceeding $20 million per MW, with some projects reaching $20 million to $25 million per MW, and short-term emergency capacity prices reaching as high as $40 million to $50 million per MW.
However, price increases have primarily occurred in short-term capacity, next-generation GPUs, large-scale clusters, and production-grade AI inference scenarios, while traditional low-priority, long-term fixed-price bare compute resources have not seen corresponding price hikes. From a profitability perspective, project-level returns for AI compute are becoming clearer, but the company-wide return on invested capital (ROIC) still requires time to validate. Nebius has for the first time disclosed more defined project payback periods, while CoreWeave has reduced GPU investment pressure through long-term contracts and asset-level financing. Nevertheless, both companies remain in a high-capital-investment phase, with depreciation and financing costs continuing to compress profit margins.
However, an increasing number of individual projects are achieving economic model闭环, indicating that the AI infrastructure business model is gradually maturing. In addition, both CoreWeave and Nebius are upgrading toward an “AI infrastructure operating system.” In the future, competition among CSPs will no longer be limited to renting GPU hours but will encompass a comprehensive service ecosystem including AI training, inference, storage, networking, model deployment, monitoring, security governance, and Agent runtime environments.
In terms of capital models, the two companies have taken different paths: Nebius leans toward a lightweight asset model, relying on capital partners to fund the construction of AI data centers while providing its own AI infrastructure operations and software capabilities; CoreWeave, on the other hand, drives a hybrid cloud model through its Omni strategy, deploying its full AI cloud platform onto customers’ own data centers and GPU resources, with a stronger emphasis on enterprise-grade and sovereign AI delivery.
Overall, the AI cloud computing industry is evolving from being a "GPU rental provider" to becoming an "AI infrastructure platform." Short-term supply constraints will continue to support compute prices, while long-term competition will shift toward capital efficiency, software capabilities, and the ability to become the operating system of the AI era.