San Francisco: OpenAI, the maker of Artificial Intelligence chatbot ChatGPT, is looking to develop its own AI chips and has even considered acquiring another company for this purpose, a media report said on Friday.
OpenAI has been discussing various options to address the shortage of expensive AI chips it relies on, but no final decision has been made, according to Reuters, citing sources.
These options have included developing its own AI chip, collaborating more closely with other chip makers, including Nvidia and also diversifying its suppliers beyond Nvidia.
OpenAI CEO Sam Altman, who has made acquiring more AI processors a primary priority for the company, is one of the main driving forces behind this decision.
He has raised concerns about the restricted availability and expensive costs of the hardware required to power OpenAI’s software, the report said.
OpenAI has been developing its generative AI capabilities on a Microsoft-built supercomputer that employs thousands of Nvidia GPUs.
The cost of running ChatGPT is prohibitively high for the company. According to an analysis from Bernstein analyst Stacy Rasgon, each enquiry costs about 4 US cents. If ChatGPT queries grew to a tenth the size of Google searches, it would require about $48.1 billion in GPUs initially and around $16 billion in chips per year to remain functioning.
Meanwhile, OpenAI is reportedly raising funds at a valuation of $80-$90 billion via the sale of existing shares. According to The Wall Street Journal, the ChatGPT developer is “talking to investors about a share sale” that would value the company between $80 billion to $90 billion, “roughly triple its level earlier this year”.
In April, OpenAI closed a more than $300 million share sale at a valuation of $29 billion.