OpenAI, the company known for its ChatGPT AI, is considering the development of its own artificial intelligence chips and has even explored potential acquisition opportunities, sources familiar with the company’s internal discussions have revealed.
While OpenAI has not yet made a final decision, recent internal talks have focused on finding solutions to address the scarcity of expensive AI chips that the company heavily relies on. These discussions have encompassed multiple options, including designing proprietary AI chips, strengthening partnerships with existing chip manufacturers such as Nvidia, and diversifying its supplier base beyond Nvidia.
According to a Reuters report, OpenAI is contemplating the development of its own AI chips to drive ChatGPT, signaling a potential departure from its partnership with Microsoft.
CEO Sam Altman has emphasised the urgent need to acquire more AI chips, given the shortage of graphics processing units (GPUs), a market largely dominated by Nvidia, which holds more than 80% of the global market share for AI-suited chips. Altman has voiced concerns about the rising costs associated with running the hardware required to support OpenAI’s projects and products.
OpenAI has been developing its generative AI technologies on a massive supercomputer provided by Microsoft, one of its key supporters, utilising 10,000 Nvidia GPUs. However, the operation costs for running ChatGPT are substantial, with each query costing approximately 4 cents, according to analysts. If ChatGPT queries were to reach a scale one-tenth that of Google’s search, it would demand an initial investment of about $48.1 billion in GPUs and approximately $16 billion worth of chips annually to maintain operations.
Read Also: What is ChatGPT and How to use this AI tool?
Exploring the development of its AI chips would place OpenAI among a select group of major tech players like Google and Amazon, who have taken control of chip design integral to their businesses.
Nonetheless, whether OpenAI proceeds with the custom chip plan remains uncertain. Building custom AI chips would represent a substantial strategic initiative and financial commitment, potentially reaching hundreds of millions of dollars in annual expenses, as per industry experts. Success is not guaranteed, even with dedicated resources.
OpenAI has contemplated the possibility of acquiring a chip company, having reached the point of conducting due diligence on a potential target, sources reveal. However, the identity of the company under consideration remains undisclosed.
Even if OpenAI decides to develop custom chips, including through an acquisition, it would likely take several years, leaving the company dependent on commercial providers like Nvidia and Advanced Micro Devices (AMD) in the interim.
Notably, some major tech companies have pursued custom chip development with mixed results. Meta, for instance, encountered challenges with its custom chip efforts, leading to the abandonment of some AI chips. Microsoft, OpenAI’s primary backer, is also working on a custom AI chip, further indicating potential divergence between the two companies.
The demand for specialized AI chips has surged since the launch of ChatGPT, as specific AI accelerators are essential for training and running advanced generative AI models. Nvidia, as one of the primary AI chip manufacturers, currently dominates this market.
Check Here Web Story on – Why Is Amazon Investing Up To 4 Billion In AI Startup Anthropic