OpenAI’s Altman Seeks Billions For AI Chip Manufacturing Venture

OpenAI chief executive Sam Altman has reportedly been courting investors in the Middle East and elsewhere to set up a network of fabrication plants for advanced AI chips.

The talks with Abu Dhabi-based AI investment group G42 and others are motivated by Altman’s belief that it is necessary to act now to ensure a sufficient supply of AI processing chips through the end of the decade, Bloomberg and other news outlets reported.

The project, which earlier reports said was code-named Tigris, would involve working with top chip manufacturers and would focus on the capital-intensive business of chip manufacturing as well as design.

Altman has repeatedly said there is insufficient AI chip supply for his company’s needs. Previous reports about the project had not indicated its focus on manufacturing.

Chip fab network

Other major tech companies, including Amazon, Google and Microsoft, have designed their own AI chips but have left manufacturing in the hands of contract firms such as market leader Taiwan Semiconductor Manufacturing Co (TSMC).

Altman reportedly held talks with potential investors in the Middle East in October before his short-lived surprise ouster the following month. He recommenced negotiations after returning to OpenAI.

The Financial Times said Altman had negotiated with investors in the United Arab Emirates including Sheikh Tahnoon bin Zayed al-Nahyan, who oversees G42 and other investment vehicles.

G42 has existing partnerships with OpenAI and Microsoft, which is OpenAI’s main backer.

TSMC talks

The newspaper also said Altman had held talks with TSMC over a partnership to fabricate AI chips.

Altman has also held talks with Microsoft about the venture and Microsoft is interested, Bloomberg said.

It is unclear whether the venture would be managed as part of OpenAI or as a separate firm. The FT said OpenAI would be the project’s primary customer.

OpenAI and other AI companies are heavily dependent on Nvidia for GPU accelerator chips, which are in heavy demand. AI services typically rely on costly and processor-intensive cloud infrastructure to carry out their tasks.

Meta chief executive Mark Zuckerberg said last week that his company was looking to accelerate its own AI efforts and was acquiring one of the world’s biggest stockpiles of Nvidia chips to further its aims.