AI datacenters priced at ~$50B per gigawatt, but build speed is the metric that matters
JOBhakdi · x · 2026-11-02
Wall Street models AI data centers in dollars per gigawatt ($50B each), but the author argues the key variable is time: SpaceX stands up a gigawatt in 6 months versus 18 for neoclouds like CoreWeave and Nebius.
A finished, contracted gigawatt generates cash and can be financed against, so faster builders can redeploy the same capital sooner — like two builders with the same loan, where the 6-month builder completes two houses in the time the 18-month builder finishes one.
The right metric is therefore dollars per gigawatt per unit of time, under which $50B buys SpaceX roughly twice the compute it buys a slower builder.
More from Infra
- OpenAI may rebrand its data centers as 'AI compute campuses' — TechCrunch AI · 2026-11-02
- $300B Louisiana project pipeline squeezes steel supply, H-beam lead times hit 9 months — MatthewChang · 2026-11-02
- Benchmark: Speedy inference engine crushes Avalon — 1.6x faster reads on Qwen3 long context — MotorNetwork380 · 2026-11-02
- vLLM serves DeepSeek-V4.1-Flash at 954K tokens/s per GPU on one GB300 NVL72 rack — bookwormengr · 2026-11-02
- Prime Mega MoE extends DeepSeek's kernel for training, up to 34% faster end-to-end MoE training — cfregly · 2026-11-02
- 2.3bpw quant update: 27B model on 12GB cards ties Qwen3.6 BF16 on LiveCode Bench at 84.3% — Brief-Tap-6616 · 2026-11-02