OpenAI is in talks with Broadcom about developing AI chip

Sam Altman is in talks with Broadcom to develop AI chips for OpenAI, reducing dependence on Nvidia GPUs.

: Sam Altman is negotiating with Broadcom and other chip designers to create AI chips for OpenAI, aiming to lessen reliance on Nvidia GPUs. OpenAI is hiring former Google employees and exploring raising funds for semiconductor factories. Although a new chip may not be ready until 2026, OpenAI is taking steps to become more self-reliant in AI development.

Sam Altman is exploring partnerships with chip designers like Broadcom to develop proprietary AI chips for OpenAI, aiming to reduce dependence on Nvidia GPUs. Discussions with Broadcom align with its experience in custom AI accelerators, similar to its collaboration with Google on tensor processing units now in their sixth generation.

OpenAI has been hiring former Google employees and considering investments in semiconductor factories as part of a broader strategy to enhance its infrastructure. This initiative aims to improve supply chains and create the necessary power and data center infrastructure to support powerful AI models.

Despite an optimistic timeline suggesting new chips might not be ready until 2026, OpenAI is exploring ways to boost self-reliance in AI innovations. The company has also opened an office in Japan and is collaborating with organizations like Khan Academy and Carnegie Mellon to push the boundaries of personalized learning through AI.

AI chipsBroadcomOpenAI