OpenAI’s CEO, Sam Altman, is presently in talks with Middle Eastern investors and contract chip manufacturers, pushing forward plans to produce home-grown Artificial Intelligence (AI) accelerators. This strategy is designed to address OpenAI’s escalating requirements for AI training and operating chips, while reducing dependency on NVIDIA.
As part of this ambitious project, Altman has negotiated with top-tier investors from the United Arab Emirates (UAE), including Sheikh Tahnoon bin Zayed Al Nahyan of Abu Dhabi, one of the most influential and wealthiest individuals. The project under discussion involves the creation and production of microchips necessary for training and operating AI models. Altman has also held discussions with TSMC for co-operation in chip production, with Samsung and Intel serving as backup contract manufacturers.
Sheikh Tahnoon, younger brother to the UAE ruler and his national security adviser, oversees a rapidly-growing business empire. He chairs Abu Dhabi Investment Authority (ADIA), with assets estimated around $800 billion, and Abu Dhabi’s sovereign wealth fund (ADQ). Additionally, he also leads the International Holding Company, UAE’s largest public company, and G42, an ambitious AI company partnering with Microsoft and OpenAI.
It is not surprising to seek investments from the East as chip development and manufacturing is a costly process demanding significant investments. Although the precise investment Altman intends to secure remains undisclosed, it is speculated that billions of dollars will be necessary to compete with NVIDIA, already nearing a market capitalization of $1.5 trillion.
The AI systems and AI chips sector has gained substantial relevance over the last year. NVIDIA’s Graphics Processing Units (GPUs) are a crucial element in sustaining this industry’s growth and are highly sought-after among tech companies both in Silicon Valley and abroad.
AI developers like OpenAI employ thousands of computation accelerators to train and operate large language AI models. Presently, OpenAI is developing a new version of its AI model, expected to launch this year as an essential upgrade from its current GPT-4 model. Each neural network update intensifies OpenAI’s dependency on chip manufacturers, especially on NVIDIA, the largest AI chip supplier.
This dependency has prompted Sam Altman to seek investors for producing their chips. Yet, it remains unclear whether this new venture will operate as a subsidiary of OpenAI or as a separate entity. Informed sources indicate that OpenAI will primarily be the main customer of the new chipmaker.