Elon Musk Says SpaceXAI’s Colossus 2 Could Run More Than 1.2 Million Nvidia GB200 And GB300 Chips By Year-End ‘If We Get Lucky’

Elon Musk said SpaceXAI’s Colossus 2 could run more than 1.2 million Nvidia Corp. (NASDAQ:NVDA) GB200 and GB300 processors by year-end if its latest expansion schedule holds.

Colossus 2 Races Toward Million-Chip Scale

Musk said Friday that Colossus 2 currently runs 110,000 GB200 chips and 440,000 GB300s. Another 220,000 GB300s will be "fully operational next week," followed by another 220,000 in November. "If we get lucky, yet another 220k GB300 by late December," Musk wrote on X.

Colossus 1 is 150k H100, 50k H200 and 30k GB200.

Colossus 2 is 110k GB200 and 440k GB300.

Another 220k GB300 will be fully operational next week and another 220k in November. If we get lucky, yet another 220k GB300 by late December.

— Elon Musk (@elonmusk) September 25, 2026

That schedule would take Colossus 2 to 1.21 million GB200 and GB300 processors, including 1.1 million GB300s. Musk said the unusual 110,000-chip increments reflect "the number of fiber optic cables that can be plugged into a central switch," highlighting networking constraints in scaling clusters.

Nvidia’s product overview for the GB300 NVL72 platform reveals that it connects 72 Blackwell Ultra GPUs in a liquid-cooled rack using fifth-generation NVLink. Nvidia says each rack can draw up to 142 kilowatts and provides 130 terabytes per second of NVLink bandwidth, showing why massive chip additions also demand cooling, networking and power infrastructure.

Power And Networking Become Critical Bottlenecks

Musk has previously said those supporting systems are becoming a bottleneck. He said this month that large AI data centers require power plants, transformers, liquid-cooling loops and complex networking, while arguing Space Exploration Technologies Corp. (NASDAQ:SPCX) moved faster by building its own power infrastructure.

That buildout, however, has faced scrutiny. Reuters reported in July that 59 natural-gas turbines had been installed to support Colossus 2 without federal clean-air permits, with environmental groups challenging the project. SpaceXAI has said it is replacing temporary generation with a 1.2-gigawatt permanent plant.

Colossus is also becoming a commercial compute business rather than merely an internal training cluster. Reuters reported separately in June that Reflection AI agreed to pay SpaceX $150 million per month for immediate access to GB300 capacity at Colossus 2.

Colossus Expands Into Commercial AI Compute

Alphabet Inc.’s (NASDAQ:GOOG) (NASDAQ:GOOGL) Google separately signed a multiyear agreement covering about 110,000 Nvidia GPUs, while Anthropic gained more than 300 megawatts at Colossus 1. The Google agreement carries a $920 million monthly price at full scale.

Photo courtesy: Shutterstock