According to reports, OpenAI, the creator of the AI chatbot ChatGPT, is actively engaged in in-depth negotiations with Broadcom, a giant in the chip industry, to jointly develop a new generation of AI chips.

This collaboration is led by OpenAI's CEO, Sam Altman, with the aim of expanding OpenAI's computing capabilities in the field of artificial intelligence. To achieve this, OpenAI has actively recruited chip design elites from Google's Tensor Processing Unit (TPU) team to explore new paths for designing chips specifically for AI servers.

Currently, technology companies are striving to follow the law of expansion, aiming to build more powerful models and relying on increasingly large computing clusters to support their operations. In this context, Altman has also been actively in contact with TSMC to explore how to improve production capacity by funding the construction of new chip factories, paving the way for OpenAI's AI chip dream.

Liz Bourgeois, a spokesperson for OpenAI, confirmed that the company is actively communicating with industry and government stakeholders to broaden access to infrastructure and ensure that the benefits of artificial intelligence technology can be widely shared in society. However, she did not provide further details on the specific negotiation with Broadcom or the chip development plan.

It is worth mentioning that OpenAI's collaboration with Broadcom is benefited from Broadcom's rich experience in the Google TPU project. The OpenAI chip team, led by former Google core member Richard Ho, clearly prefers to work with this American company.

Altman's blueprint goes beyond chip design. He has been discussing with external investment forces to establish a new company dedicated to investing in real estate, power infrastructure, data centers, and specialized AI chip servers in key areas. In the future, OpenAI will lease these servers to support its massive AI computing needs.

This strategic layout may give OpenAI a more advantageous position in negotiations with AI chip giants such as NVIDIA. NVIDIA has made a fortune with its focus on AI graphics processors (GPUs), and OpenAI's self-developed chip plan undoubtedly poses potential competition to it. However, this may also make the relationship between OpenAI and this important supplier delicate and complex.

Currently, the proposed AI chip development and supporting infrastructure construction projects are still in the preliminary conceptual stage, and the chip design process has not yet officially started. Mass production is expected to be delayed until 2026 or later. However, the intensive discussions with industry partners such as Broadcom have fully demonstrated OpenAI's firm determination and relentless pursuit of realizing this grand vision.