OpenAI Targets $600B Spend by 2030
OpenAI has reset its long-term spending expectations, now targeting approximately $600 billion by 2030. The figure reflects the massive anticipated scale of capital required for computing infrastructure and research to achieve its goals in artificial general intelligence. This level of investment underscores the immense resource requirements for companies competing at the frontier of AI development.
- This revised spending target is a significant reduction from the $1.4 trillion figure CEO Sam Altman had previously mentioned. The new figure is intended to more closely align with the company's projected revenue of over $280 billion by 2030. - A major driver of these costs is the need for specialized AI chips and the construction of data centers to house them. A single high-end AI server rack can cost over $500,000, and these servers consume up to 10 times more power than traditional ones. - To address the massive need for semiconductors, Sam Altman has been in discussions with global investors to fund a network of AI chip fabrication plants. Earlier, more ambitious reports mentioned he was seeking $5 to $7 trillion for this effort, a figure that dwarfs the entire global semiconductor industry's sales. - The cost of training frontier AI models has been growing at a rate of 2.4 times per year since 2016. While training GPT-3 in 2020 cost an estimated $2-4 million, the technical cost for GPT-4 rose to between $41 million and $78 million. - OpenAI is in the process of closing a funding round that could exceed $100 billion, with significant investments expected from strategic partners like Nvidia, SoftBank, and Amazon. This funding would support a pre-money valuation of around $730 billion. - The global semiconductor market is projected to reach nearly $1 trillion in annual sales by 2030, largely driven by the demand for AI chips. This highlights the immense market that OpenAI and its competitors are creating for chip manufacturers. - In 2025, OpenAI's revenue reached $13.1 billion, surpassing its internal goal of $10 billion. During the same period, the company's spending was $8 billion. - Expenses for running AI models, known as inference costs, are a significant and growing part of the financial picture. For OpenAI, these costs reportedly quadrupled in 2025, which impacted the company's gross margins.
