Elon Musk has set an aggressive target for SpaceX, stating he is "cautiously optimistic" that the company will develop an AI model on par with GPT-6 or Fable within the next two to three months. In posts on X, the billionaire suggested that if the current pace of development continues, SpaceX could lead the AI field in approximately six months.
The Power of Colossus
A central pillar of Musk's strategy is raw computational power. SpaceX is investing heavily in its Colossus systems, which leverage advanced Nvidia processors. According to data released by Musk:
- Colossus 1 consists of 150,000 H100s, 50,000 H200s, and 30,000 GB200s.
- Colossus 2 features 110,000 GB200s and 440.000 GB300s.
Musk announced an ambitious expansion timeline, with 220,000 additional GB300 units planned for deployment next week, followed by another 220,000 in November and a final batch by late December. He maintains that SpaceX's expertise in complex engineering projects provides a competitive edge in rapidly scaling large-scale hardware infrastructure.
Expansion into Enterprise AI
Beyond hardware, SpaceX is strengthening its position through the acquisition of Cursor, an AI platform for developers. This move signals a shift toward enterprise AI, moving beyond the company's traditional focus on space launches and Starlink. According to JPMorgan estimates, Cursor generates approximately $4 billion in annual recurring revenue, with 75% derived from corporate clients.
Despite Musk's optimism, these projections remain personal estimates and have yet to be verified by independent performance benchmarks or third-party audits.
