OpenAI plans to invest $1.15 trillion in hardware and cloud infrastructure from 2025 to 2035. The spending will be divided among seven major vendors: Broadcom, Oracle, Microsoft, Nvidia, AMD, Amazon AWS, and CoreWeave. Broadcom receives the largest share at $350 billion, followed by Oracle at $300 billion and Microsoft at $250 billion. The spending is expected to ramp up significantly, with annual compute spending projected to rise from $6 billion in 2025 to $295 billion by 2030. In terms of revenue, OpenAI anticipates a gross profit margin starting at 48% in 2025 and reaching 70% by 2029. To sustain the planned infrastructure spending, OpenAI would need to generate substantial revenue growth, climbing from approximately $10 billion in 2024 to about $577 billion by 2029. This trajectory parallels the expected revenue size of Google in that same year, illustrating the scale of OpenAI's ambitions. The article outlines the estimated spending timelines and contract structures for each vendor. For example, OpenAI's deal with Oracle includes $60 billion annually for five years, while Broadcom's contract is based on deploying 10 gigawatts of custom AI accelerators. The estimates reflect a gradual ramp-up, with major spending beginning after 2027 as infrastructure becomes operational. The complexity of these contracts involves various payment structures, making precise modeling challenging, yet the projected growth rates signal a rapid escalation in OpenAI's operational capabilities.