SpaceXAI could push its Colossus AI infrastructure to roughly 1.44 million NVIDIA GPUs by the end of 2026 if its latest deployment schedule holds.

Elon Musk said Colossus 1 currently contains 150,000 NVIDIA H100 GPUs, 50,000 H200s and 30,000 GB200s. Colossus 2 contains 110,000 GB200s and 440,000 GB300s.

The next phase is considerably larger. Musk said another 220,000 GB300 GPUs should become fully operational next week, followed by 220,000 more in November. A third 220,000-GPU batch could become operational in late December “if we get lucky.” If all three batches arrive on schedule, the combined Colossus fleet would reach about 1.44 million GPUs.

Musk detailed the hardware rollout directly on X:

The scale makes power delivery nearly as important as GPU supply. Reporting on the expansion points to a 1.2-gigawatt power project intended to support the growing compute footprint as more systems come online.

The newest expansion is centered on NVIDIA’s GB300 platform. Reaching the full 1.44-million-GPU figure depends on SpaceXAI completing all three planned deployment waves, so the year-end number remains a target rather than an already-operational total.

Sources: Tom’s Hardware and Elon Musk’s X post.

BitcoinVersus.Tech Editor’s Note:

We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate here: 3C9o19EH5HSiwEPyCTmEKzxhNCbo2X6TTb

BitcoinVersus.tech is not a financial advisor. This media platform reports on financial subjects purely for informational purposes.