The retained GPTOnline cover is an editorial AI-chip graphic with a headline about the October 2023 report. It is not a photograph of an OpenAI chip or evidence of a manufacturing announcement.
In October 2023, Reuters reported that OpenAI, the company behind ChatGPT, was exploring custom AI chips and had evaluated a potential acquisition target. The report cited people familiar with the discussions; it did not describe a final decision or an official chip-launch announcement.
The decision to proceed was still unresolved at that time, as per internal discussions described to Reuters in 2023. However, since at least 2022, OpenAI had been considering various strategies to tackle the scarcity of high-cost AI chips it heavily relies on. These deliberations had included the possibility of developing its own custom AI chip, strengthening ties with other chipmakers including Nvidia, and expanding its supplier base beyond Nvidia. Developing a custom design would not, by itself, establish a plan to own or operate a semiconductor fabrication plant.
OpenAI declined to comment on the October 2023 report. However, CEO Sam Altman has been vocal about the need for more AI chips, making it a priority for the company. He has expressed discontent about the dearth of graphics processing units (GPUs), a market in which Reuters reported Nvidia held more than 80% of the relevant AI-chip segment at the time. This is a historical estimate, not a current market-share figure.
Altman’s push for more chips is driven by two primary concerns: the shortage of advanced processors that fuel OpenAI’s software, and the hefty costs associated with operating the hardware necessary to power its products and efforts.
Microsoft announced a supercomputer built for OpenAI in May 2020 with 10,000 GPUs. That figure describes the system announced at the time, not OpenAI’s total or current computing infrastructure.
Reuters cited a historical estimate by Bernstein analyst Stacy Rasgon of roughly 4 cents per ChatGPT query. The same analysis projected about $48.1 billion in initial GPU spending and $16 billion in annual chip spending if query volume reached one-tenth of Google Search’s scale. These were estimates for a hypothetical scenario, not disclosed OpenAI expenditure or current per-query prices.
The move to develop its own AI chips would position OpenAI among a select group of major tech companies, including Google and Amazon, that have sought to gain control over the design of chips integral to their operations.
At the time of the report, it remained unclear whether OpenAI would proceed with plans for a custom chip. Such a move would represent a significant strategic undertaking and a substantial investment potentially running into hundreds of millions of dollars per year, as per industry insiders. Even dedicating resources to this endeavor would not guarantee success.
Acquiring a chip company could expedite the process of creating OpenAI’s proprietary chip, similar to Amazon’s acquisition of Annapurna Labs in 2015.
OpenAI had considered this route to the extent that it performed due diligence on a potential acquisition target, according to a source close to its plans. The identity of the company OpenAI was considering purchasing was not identified in the report.
The report suggested that, if OpenAI pursued a custom chip and possible acquisition, the work could span several years, leaving the company reliant on commercial providers like Nvidia and Advanced Micro Devices in the interim.
Other major tech firms have attempted to build their own processors for years, often with limited success. Reuters described setbacks in Meta’s custom chip project at that time, including the abandonment of some designs. The report said the Facebook owner was then focusing on a newer chip designed to cater to all kinds of AI work.
The report also cited The Information’s account that Microsoft was developing a custom AI chip that OpenAI was testing. That report did not establish that the companies’ partnership was ending.
The demand for specialized AI chips has skyrocketed since the launch of ChatGPT in November 2022. These specific chips, or AI accelerators, are crucial for training and running the latest generative AI technology. The article described Nvidia as the leading supplier in that historical market; it did not provide a current comparison of all AI-chip vendors.
