CoreWeave is leveraging AI to significantly enhance its revenue through substantial investments in AI infrastructure and data center capacity, targeting $20 billion to $23 billion in spending this year to meet the escalating demand from major clients like Microsoft.
This strategy is underscored by a $25.9 billion revenue backlog, driven largely by a five-year deal with ChatGPT maker OpenAI worth $11.2 billion, where CoreWeave provides AI infrastructure in exchange for a stake in OpenAI.
Despite the heavy capital outlay, CoreWeave’s CEO Mike Intrator, in the earnings report, emphasizes that the company’s unique debt and capital structuring approach ensures profitable returns by fully loading infrastructure costs into repayments.
While Microsoft’s demand is anticipated to be stable for the next three years, CoreWeave is positioned as overflow capacity, indicating potential revenue fluctuations beyond that period, Reuters news report said. Additionally, CoreWeave is mitigating risks associated with U.S.-China trade tensions by diversifying its supply chains, aiming to secure margins and sustain profitability amidst intense competition in the AI infrastructure sector.
CoreWeave reported a strong start to 2025, driven by substantial revenue growth and strategic customer acquisitions, including a $11.2 billion deal with OpenAI, positioning the company as a key AI infrastructure provider.
CoreWeave’s revenue has shown impressive growth over recent quarters, increasing from $189 million in Q1 2024 to $395 million in Q2 2024, $584 million in Q3 2024, $747 million in Q4 2024, and reaching $982 million in Q1 2025. This marks a 420 percent year-over-year increase, driven by strong demand for its cloud platform and long-term committed contracts that provide strong revenue visibility and support strategic capital investments.
The first quarter of CoreWeave ended with a $25.9 billion revenue backlog, comprising $14.7 billion in remaining performance obligations (RPO) and an additional $11.2 billion under committed contracts. Notable client wins included partnerships with AI labs, hyperscalers, and enterprises, most prominently IBM, to supply compute capacity for its Granite models.
CoreWeave also achieved significant operational scaling, increasing compute capacity to 420 MW of active power and securing 1.6 GW of contracted power.
The company strengthened its AI Hyperscaler leadership with record-breaking MLPerf Inference v5.0 results using NVIDIA GB200 Grace Blackwell Superchips, earning a Platinum rating from SemiAnalysis’s ClusterMAX system.
CoreWeave has launched new products, including CoreWeave AI Object Storage (CAIOS) for high-performance data access, on-demand cluster scaling via Kubernetes, and enhanced support for NVIDIA AI Enterprise software, further expanding its AI cloud platform capabilities.
CoreWeave has raised $1.4 billion through its IPO, bringing total capital raised to $17.2 billion, underscoring its aggressive strategy to drive the next generation of AI-driven cloud infrastructure.
Baburajan Kizhakedath
