OpenAI最新芯片策略或降低ChatGPT最大开支

OpenAI has taken what may be its most significant step yet toward reducing its dependence on Nvidia, a move that could eventually lower the cost of running ChatGPT and other artificial intelligence services. The effort is not about replacing Nvidia overnight. Instead, it reflects a broader shift taking place across the AI industry, where the world’s largest model developers are increasingly looking beyond general-purpose AI accelerators and investing in hardware tailored to their own workloads. For OpenAI, the economics are difficult to ignore. Every conversation with ChatGPT requires computing power, and as hundreds of millions of people use AI tools daily, infrastructure has become one of the company’s largest operating expenses. …