OpenAI, 2026’da Kendi Yapay Zeka Çipini Üretecek

First Chip in 2026

According to a report by Financial Times, the chip developed under this project will not be sold to external customers. It will be used exclusively for training OpenAI’s AI models, such as ChatGPT, and for operating its platforms. This move is seen as part of the company’s strategy to reduce its reliance on Nvidia and cut costs.

OpenAI’s chip plans are not actually new. Last year, it was reported that the company was working on a custom chip in collaboration with Broadcom and TSMC. Additionally, OpenAI is also utilizing AMD chips alongside Nvidia to meet the increasing demand for infrastructure.

Meanwhile, Broadcom CEO Hock Tan announced that they received an AI infrastructure order exceeding $10 billion from an undisclosed new customer in the last quarter and expect significant revenue growth in this area for the 2026 fiscal year. Analysts note that this customer could potentially be OpenAI.

OpenAI’s move parallels the efforts of Google, Amazon, and Meta to develop their own AI chips. Major tech companies are turning to their own hardware solutions to manage the rapidly increasing costs of AI training and operation.