Report: SpaceX Orders Million-Level GB300 GPUs, Nearing 2GW
ns123abc · x · 2026-07-20
According to reports, SpaceX just ordered about 1 million new GB300 GPUs, pushing its Colossus supercomputer cluster's compute scale to nearly 2GW.
Foxconn is reportedly named as the manufacturing partner. Specific procurement details include:
- Over 13,000 racks, translating to about 936,000 GPUs
- Cost per rack is roughly $4 million, with a total order value of about $52 billion
- Expected delivery by the end of 2026
Related event: Rumor: SpaceX Orders $52B Worth of GB300 Servers from Foxconn(8 posts)→
More from Infra
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27
- Gary Marcus: With Q3 data added, Anthropic's compute-efficiency growth shows deceleration — GaryMarcus · 2026-10-27
- Inference is just Kubernetes with a multi-trillion-dollar market, and the communities overlap — matt_slotnick · 2026-10-27
- vLLM eyes Rust migration as inference engines race to rewrite the stack — AccBalanced · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27