OpenAI is entering the closing stages of designing its prolonged-rumored AI processor with the goal of lowering the firm’s dependence on Nvidia hardware, in line with a Reuters file released Monday. The ChatGPT creator plans to send its chip designs to Taiwan Semiconductor Manufacturing Co. (TSMC) for fabrication for the duration of the following few months, nonetheless the chip has no longer but been formally announced.
The OpenAI chip’s full capabilities, technical crucial components, and trusty timeline are peaceable unknown, nonetheless the firm reportedly intends to iterate on the form and affords a boost to it over time, giving it leverage in negotiations with chip suppliers—and doubtlessly granting the firm future independence with a chip kind it controls outright.
Prior to now, we now comprise considered completely different tech companies, such as MicrosoftAmazonGoogleand Metamanufacture their own AI acceleration chips for reasons that range from mark reduction to relieving shortages of AI chips equipped by Nvidia, which enjoys a attain-market monopoly on high-powered GPUs (such as the Blackwell series) for data heart exercise.
In October 2023, we coated a file about OpenAI’s design to fabricate its own AI accelerator chips for identical reasons, so OpenAI’s custom chip mission has been in the works for a whereas. In early 2024, OpenAI CEO Sam Altman also began spending appreciable time touring spherical the arena looking out out for to lift up to a reported $7 trillion to expand world chip fabrication capability.
