OpenAI’s Chief Executive Officer (CEO), Sam Altman, aims to secure billions for an AI chip venture, intending to use the funds to establish a global network of factories for chip fabrication.
The effort involves collaboration with undisclosed top chip manufacturers, as reported by Bloomberg.
Addressing a crucial aspect of running AI models, such as ChatGPT and DALL-E, the report highlights the challenge of having sufficient chips for computations.
Nvidia’s value exceeded $1 trillion last year, driven in part by its virtual monopoly, with GPT-4, Gemini, Llama 2, and other models heavily relying on its popular H100 GPUs.
The competition to manufacture high-powered chips for complex AI systems has intensified, given the limited number of fabs capable of producing such chips.
“Mr Altman and others are bidding for capacity years in advance to meet the demand,” the report noted. “To compete with industry giants like Apple, deep-pocketed investors are essential, prompting talks with SoftBank Group and G42, an Abu Dhabi-based AI holding company, to raise funds for Altman’s project.”
Notably, Microsoft and Amazon have ventured into making their own AI chips, with Google using its DeepMind AI to design processors like Tensor Processing Units (TPU).
Meta’s CEO, Mark Zuckerberg, revealed plans to own over 340,000 of Nvidia’s H100 GPUs by the end of the year as Meta pursues artificial general intelligence (AGI) development.
Nvidia is already gearing up with its next-generation GH200 Grace Hopper chips, while competitors like AMD, Qualcomm, and Intel have launched processors designed to power AI models on various devices, including laptops and phones.