OpenAI, the organization responsible for ChatGPT, is exploring the possibility of creating its own artificial intelligence chips. According to sources familiar with the company’s plans, they have even considered the acquisition of a potential target in this regard. However, as of recent internal discussions, OpenAI has not made a final decision to proceed. They have been actively exploring options to address the shortage of expensive AI chips they rely on, including the potential development of their own AI chip, closer collaboration with chip manufacturers like Nvidia, and diversifying their suppliers beyond Nvidia.
OpenAI’s CEO, Sam Altman, has emphasized the acquisition of more AI chips as a top priority for the company. He has expressed concerns about the scarcity of graphics processing units (GPUs), which are essential for AI applications and are predominantly supplied by Nvidia. Altman’s concerns are driven by both a shortage of advanced processors needed for OpenAI’s software and the high costs associated with operating the necessary hardware.
OpenAI has been utilizing a massive supercomputer, powered by 10,000 Nvidia GPUs, for its generative AI technologies since 2020. Running ChatGPT comes at a substantial cost, estimated at around 4 cents per query. If ChatGPT queries were to reach a scale similar to Google search, it would require a substantial investment in GPUs.
Should OpenAI decide to develop its own AI chips, it would join a select group of major tech companies, such as Google and Amazon, that have pursued chip design to align with their business needs. However, it remains uncertain whether OpenAI will proceed with this plan, as it would be a significant strategic initiative and entail substantial financial investments, with no guaranteed success. Alternatively, OpenAI could explore the acquisition of a chip company, similar to Amazon’s acquisition of Annapurna Labs in 2015, which could expedite the development of their custom chip.
While OpenAI’s efforts in custom chip development may take several years, they will likely continue to rely on commercial providers like Nvidia and Advanced Micro Devices (AMD) in the interim. Notably, other major tech companies, such as Meta (formerly Facebook), have faced challenges in developing their own custom chips.
The demand for specialized AI chips has surged since the introduction of ChatGPT, as these chips are essential for training and running advanced generative AI models. Nvidia currently dominates the market as one of the primary suppliers of AI chips.