In an effort to control more of its own destiny, ChatGPT maker OpenAI is reportedly focused on entering the AI chip market either through acquisition or buy building out its own manufacturing arm.
Sam Altman, CEO of OpenAI, has stated that the manufacture of AI chips is a core strategic goal for OpenAI.
AI chips are expensive to produce with few companies able to manufacture them.
Nvidia currently is the main provider of these components, controlling roughly 80% of the market.
The Nvidia chip runs a cool $30,000 and it is estimated that one of the processing centers that Microsoft built for OpenAI’s ChatGPT training processing uses 10,000 in one facility alone.
The gold rush to general AI intelligence is what is pushing these huge expansions in hardware computing power, and the main blocker to scaling these computer centers even further is the costs and available quantity of these AI chips.
AI chips are optimized for tasks associated with machine learning and data analysis, making them better suited for these activities compared to traditional chips.
They can be tens or even thousands of times faster and more efficient than traditional Central Processing Units and are designed to handle a vast amount of data and conduct numerous calculations simultaneously, catering to the data-intensive nature of AI applications.
It is unclear if OpenAI will pursue a buy or build strategy, but many companies are jumping into the race to compete with Nvidia including, Taiwan Semiconductor, Samsung, IBM, Intel, and big tech companies like Google Alphabet, and Amazon.
Semiconductor giant AMD recently announced its plans to enter the market later this year.