OpenAI, the innovative company behind the widely popular ChatGPT, has announced a significant collaboration with chipmaker Broadcom to design its proprietary artificial intelligence computer chips. While the financial details of this agreement remain undisclosed, the two California-based firms revealed plans to commence the deployment of these new racks of customized “AI accelerators” as early as late next year. This strategic move represents the latest in a series of major partnerships forged by OpenAI with entities crucial to the development and housing of AI infrastructure, including chip manufacturers and data center providers.

The initiative to develop custom chips is not new for OpenAI; CEO Sam Altman confirmed that work on these accelerators began over a year ago. Altman articulated that the creation of their own accelerators contributes to a more extensive ecosystem of partners collectively striving to build the necessary capacity to advance the frontiers of AI, ultimately aiming to provide benefits to all humanity. Echoing this sentiment, Broadcom CEO Hock Tan expressed enthusiasm for the collaboration, stating their excitement to co-develop and deploy “10 gigawatts of next generation accelerators and network systems to pave the way for the future of AI.” Following the Monday morning announcement, Broadcom’s shares experienced a notable surge, climbing more than 9%.

This partnership with Broadcom is part of a broader, aggressive strategy by OpenAI to secure and enhance its AI infrastructure. In recent weeks, the AI startup has also established agreements with other prominent chipmakers, including Nvidia and AMD, to procure specialized chips vital for running its sophisticated AI systems. Beyond chip supply, OpenAI has entered into substantial deals with companies such as Oracle and CoreWeave, which are instrumental in developing and operating the massive data centers where these powerful AI chips are housed.

The nature of many of these deals, characterized by circular financing where companies both invest in OpenAI and simultaneously supply it with essential technology, has sparked discussions and concerns within financial institutions regarding a potential “AI bubble.” Despite currently not turning a profit, OpenAI demonstrates substantial market penetration, with its flagship chatbot, ChatGPT, reportedly attracting over 800 million weekly users. The company’s relentless pursuit of robust AI infrastructure underscores its commitment to continuous innovation and expansion within the rapidly evolving field of artificial intelligence.