Musk Says Colossus 2 Plans to Add Nvidia AI Chips by End of 2026

Elon Musk said SpaceXAI (NASDAQ:SPCX) plans to increase the number of advanced Nvidia (NASDAQ:NVDA) artificial intelligence chips used at its Colossus 2 data centre in Memphis, Tennessee, by the end of 2026.

In a post on X early Friday, Musk said the facility has used 110,000 Nvidia GB200 chips and 440,000 GB300 chips.

Musk, chief executive of SpaceX, of which SpaceXAI is a unit, said an additional 220,000 GB300 chips are expected to come online by the following week, with another 220,000 planned for November.

“If we get lucky, yet another 220k GB300 by late December,” Musk said in response to a post concerning xAI’s computing capacity.

Colossus Computing Capacity Expands

Musk has described the Colossus facility as the “world’s largest AI supercomputer.”

Colossus 1 was initially built in 2024 to provide computing capacity for training Grok, xAI’s artificial intelligence model.

The complex has subsequently expanded to provide computing resources for other projects.

According to the supplied information, xAI began renting computing capacity from the facility to Anthropic and Alphabet’s Google earlier in 2026.

The planned additions outlined by Musk would substantially increase the number of Nvidia GB300 processors operating at Colossus 2 if implemented according to the stated schedule.

Power and Capacity Requirements Not Specified

Musk did not provide details on whether Colossus 2 currently has sufficient available power and infrastructure capacity to operate all of the additional Nvidia processors.

Data centres supporting large-scale artificial intelligence workloads require computing infrastructure as well as sufficient power and associated systems to operate the hardware.

The timing of the planned expansion therefore remains based on Musk’s stated schedule, with no additional information provided regarding the facility’s available power capacity.

SpaceX stock price

Nvidia stock price