Foxconn wins $52B SpaceX AI server mega-contract

If the reported numbers are even directionally right, this is one of the largest single AI hardware orders ever placed. A $52 billion contract for 13,000+ GB300 server racks would put SpaceX’s Colossus complex at the bleeding edge of global AI compute, rivaling or surpassing the largest known datacenter clusters dedicated to model training and inference. For the race to AGI, more GPU‑dense clusters directly translate into more and larger experiments: bigger models, longer context, more reinforcement learning and more ambitious multi‑agent runs.

It also tells us that Elon Musk is serious about positioning SpaceX — not just xAI — as a compute superpower. Colossus 1 already hosts over 220,000 Nvidia GPUs; Colossus 2 is past 550,000, and this order pushes further toward million‑GPU scale.([aihot.virxact.com](https://aihot.virxact.com/items/cmrskaxl509twbiwmh4gptwaq)) That kind of capacity gives one actor outsized influence over what frontier experiments are feasible, especially if paired with proprietary data streams from satellites, vehicles and communications.

For Foxconn, becoming the assembler of choice for AI superclusters marks a strategic pivot from smartphones to AI infrastructure. It fragments the traditional server OEM oligopoly and could compress margins at Dell and Supermicro, while making it easier for deep‑pocketed buyers to stand up custom AI factories at scale.