OpenAI Slashes Compute Spending Plans
OpenAI has reportedly cut its planned spending on compute infrastructure by more than half, now targeting $600 billion instead of a previously discussed $1.4 trillion. The adjustment signals a re-evaluation of capital efficiency and cost structures as the AI sector matures. The news follows layoffs at crypto exchange Gemini, which is restructuring and cutting 25% of its staff.
- The revised $600 billion spending target through 2030 is now more closely aligned with OpenAI's projected annual revenue, which is expected to surpass $280 billion by the same year. For context, the company reportedly generated $13.1 billion in revenue in 2025 while spending $8 billion. - This adjustment follows investor concerns about the financial viability of the initial $1.4 trillion figure, a number that would have been a significant portion of the entire semiconductor industry's projected global revenue. - Concurrent with these revised spending plans, CEO Sam Altman has been in discussions with investors to raise a staggering $5 to $7 trillion for a separate, immensely ambitious project. This initiative aims to build a global network of semiconductor fabrication plants to address the ongoing global shortage of specialized AI chips. - To decrease its reliance on any single hardware provider, OpenAI is actively diversifying its chip suppliers. The company has entered into a multi-year partnership with Broadcom to co-develop custom AI accelerators. Deployments of these custom chips are scheduled to begin in the second half of 2026. - This move into designing its own chips follows a trend set by other tech giants like Google with its Tensor Processing Units (TPUs) and Amazon with its Trainium and Inferentia chips. Owning the chip design allows for greater optimization of hardware for specific AI models, which can lead to improved performance and efficiency. - The competitive landscape for AI chips is heating up, with established players like AMD and Intel vying for market share against the current leader, Nvidia. This increasing competition could lead to more options and potentially lower costs for companies building AI products. - By getting more involved in the hardware ecosystem, from chip design to manufacturing, OpenAI aims to build a more resilient and cost-effective supply chain for the massive computational resources its advanced models require. This strategy is seen as critical for long-term growth and leadership in the AI sector.
