Sequoia's Maguire models SpaceX taking inference compute share as neoclouds struggle
shaunmmaguire · x · 2026-11-07
Sequoia investor Shaun Maguire published key assumptions behind his SpaceX model: hyperscalers are delaying big projects while neoclouds "deeply struggle," ranging from severe delays to outright bankruptcy. His scenario: inference demand grows 3x by end of 2027, SpaceX sharply increases its share of inference compute supply, and some neoclouds are forced to sell hardware to SpaceX at a discount starting around 2028. This keeps SpaceX's rev/GW high while its CapEx/GW slightly declines. He also notes SpaceX should have 4 operational Starship launch pads by end of 2026, and says he's studied the company for 10 years.
More from Infra
- Tesla targets chip fab inefficiency: wafers spend as little as 20% of time in useful processing — XFreeze · 2026-11-07
- Redditor runs Qwen Flash Next locally at 24 tok/s on 64GB MacBook via Strata SSD offloading — New-Pressure-6932 · 2026-11-07
- Reactor co-founder publishes 'The real-time manifesto' on serving real-time AI at scale — smtabatabaie · 2026-11-07
- Clairvoyance's new hivemind feature lets you share your GPU with anyone — draginol · 2026-11-07
- zml/llmd inference server runs on NVIDIA, AMD, Intel, TPU and Metal with images as small as 142MB — siwu · 2026-11-07
- Musk teases 'Starmind computer' details next year, promising it will 'blow people's minds' — ns123abc · 2026-11-07