xAI founder aiming for 1 billion GPUs to scale Generative AI operations

xAI founder, Elon Musk has confirmed that his company is planning to scale operational capacity with at least 1 billion GPUs.

A recent report suggested that xAI has planned to improve the operational capacity at the Memphis Tennessee Supercomputer with 1 million GPUs. This will be a 10 times improvement of the current 100,000 GPUs stacked in server racks at the Supercomputer cluster.

“Nope, at least 1 billion GPUs!” reacted Musk in this report of expansion.

The billionaire has shared no further information on this matter and it’s hard to unfold his intention because of the pace of xAI’s development.

However, in the comment section of his X post, Musk also gave a node to a trillion GPUs. That’s a long shot but we’ll stick to 1 billion GPUs for the time being.

Over the past 10 months, the new AI company has made remarkable progress in Large Language Models (LLM) and generative AI. It launched Grok, its AI chatbot and already seeded two major versions with new features including generative image and content analysis.

Last month, xAI raised $6 billion to increase operations with additional GPUs and the Memphis facility. That said, it’s preparing future products including Grok 3, which will improve the feature and response speed.

This new development comes as xAI continues to evolve as a competitor to OpenAI’s Chatbot, Google’s Gemini, Anthtorpic’s Claude, and more. It has a unique edge compared to its competitors with X’s (formerly Twitter) integration, offering its services to millions of social media users.

The AI company has completed the current Colossus supercomputer in 122 days, which is by far the fastest AI server build. The additional 1 million GPUs should be installed within the next year. Also, achieving a 1 billion GPU milestone will take more time than that.