OpenAI expands infra via partnership with AI chipmaker Cerebras
OpenAI’s partnership with Cerebras will utilise specialised high-speed chips to accelerate complex AI reasoning and model outputs.
ChatGPT creator OpenAI has announced a multi-year partnership with the artificial intelligence (AI) chipmaker Cerebras to enhance its computing infrastructure.
This agreement will see Cerebras deliver 750 megawatts of high-speed compute power to the OpenAI platform between now and 2028. Reuters has reported that the deal is worth over $10 billion over the life of the contract.
The additional capacity will be introduced in phases, starting this year, to help the ChatGPT maker meet the soaring demand for its services.
The partnership focuses on inference, which is the process where an AI model generates a response to a user query. Cerebras specialises in purpose-built AI hardware that differs from traditional GPU systems by using a single giant chip to combine compute and memory. This design aims to eliminate the bottlenecks that often slow down complex tasks such as code generation or image creation.
“Just as broadband transformed the internet, real-time inference will transform AI, enabling entirely new ways to build and interact with AI models,” said Andrew Feldman, Co-founder and CEO of Cerebras.
By reducing wait times, OpenAI hopes to encourage users to run higher-value workloads and engage with more natural, conversational interfaces.
Founded in 2015, Cerebras has emerged as a contender to established players like NVIDIA. The company has faced a complicated journey toward the public markets, having postponed its initial public offering several times since 2024.
Interestingly, OpenAI chief executive Sam Altman was an early investor in Cerebras, and his organisation previously considered acquiring the chipmaker entirely.
For OpenAI, the deal is part of a broader strategy to build a resilient portfolio of diverse hardware solutions.
“Cerebras adds a dedicated low-latency inference solution to our platform. That means faster responses, more natural interactions, and a stronger foundation to scale real-time AI to many more people,” remarked Sachin Katti of OpenAI.
This expansion comes at a time when OpenAI is reportedly laying the groundwork for its own massive public listing. While the industry continues to attract eye-watering valuations, some experts have cautioned that the heavy spending on hardware might signal a bubble reminiscent of the dot-com era.


