Bloomberg's latest report states that Sam Altman, CEO of OpenAI, is raising billions of dollars for an artificial intelligence chip company. The purpose is to use this funding to develop a "factory network" that will be spread globally and collaborate with unnamed "top chip manufacturers".

One of the main costs and limiting factors in running artificial intelligence models is having enough chips to process the calculations behind robots like ChatGPT or DALL-E, which generate responses and images. Last year, Nvidia's market value exceeded $1 trillion, partly due to its virtual monopoly position, as models like GPT-4, Gemini, Llama 2, and others heavily rely on its popular H100 GPU.

Therefore, the competition to manufacture more high-powered chips for running complex artificial intelligence systems is intensifying. The number of wafer fabs capable of producing high-end chips is limited, prompting Altman or anyone else to compete for capacity years before it is needed in order to produce new chips. Competing with companies like Apple requires deep-pocketed investors to bear the costs that OpenAI, as a non-profit organization, still cannot afford. It is reported that SoftBank Group and Abu Dhabi AI Holdings company G42 have been in talks to raise funds for Altman's project.

Other companies developing artificial intelligence models have also ventured into manufacturing their own chips. Microsoft, one of OpenAI's investors, announced in November that it has produced its first custom AI chip for training models. Shortly after, Amazon announced the launch of its new Trainium chip. Google's chip design team is using its DeepMind AI running on Google Cloud servers to design artificial intelligence processors, such as its Tensor Processing Units (TPUs).

AWS, Azure, and Google are also using Nvidia's H100 processor. Meta CEO Mark Zuckerberg stated that "by the end of this year, Meta will have over 340,000 Nvidia H100 GPUs" as the company strives to develop Artificial General Intelligence (AGI).

Nvidia has already released the next-generation GH200 Grace Hopper chip to expand its dominant position in the field, while competitors AMD, Qualcomm, and Intel have also introduced processors aimed at powering artificial intelligence models on laptops, smartphones, and other devices.