JAKARTA - The US ruler, Elon Musk, recently revealed to investors that his artificial intelligence startup (AI), xAI, plans to build a supercomputer to support its latest version of its AI chatbot, Grok. This was reported by The Information on Saturday, May 24, citing a company presentation to investors.

Musk stated that he wanted the proposed supercomputer to be operational by fall 2025. He also added that xAI could partner with Oracle to develop the giant computer.

The xAI itself could not immediately be reached for comment. Oracle also did not respond to requests for comment from the media.

When completed, Nvidia's flagship graphics processing unit (GPU) of Nvidia will be at least four times larger than the largest GPU cluster today, The Information reports citing Musk from a presentation made to investors in May.

The strong H100 GPU family from Nvidia dominates the data center chip market for AI but can be difficult to get because of high demand.

Musk founded xAI last year as a challenger for Alphabet's Microsoft and Google-backed OpenAI. Musk also co-founded OpenAI.

Earlier this year, Musk said that Grok 2 model training required around 20,000 Nvidia H100 GPUs, adding that the Grok 3 model and beyond would require 100,000 Nvidia H100 chips.


The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)