GB300 Rents for $12-15/hr as Inference Supply Outstrips a Thin Buyer Market
teodorio · x · 2026-11-10
Morph's skeptrune argues the AI infrastructure boom has an overlooked problem: not enough customers for all the inference capacity being built. GB300 GPUs rent for just $12-15/hour even though serving DeepSeek could theoretically generate $34/hour.
Key points:
- All of OpenRouter probably amounts to at most 16 racks of GB300, against tens of thousands of racks available worldwide.
- If you build a 20MW data center to sell tokens, fewer than 1,000 companies globally would actually buy from you. "You can't sell a gigawatt of compute as inference. There's not enough demand."
- The reposter adds that beyond coding, the industry still lacks persistent large-scale use cases — and even when found, change management remains a huge barrier; promise lies mostly in downstream use cases.
More from Infra
- simdjson disables 256-bit LASX on Loongson 3A6000: 128-bit LSX is 6.7% faster — lemire · 2026-11-10
- Vera launches gyms, benchmarks and agents for cloud infrastructure operations — tianyin_xu · 2026-11-10
- NVIDIA paper: masking ~10 'intrusion heads' recovers 57% of distractor accuracy loss — rohanpaul_ai · 2026-11-10
- Memory pricing itself out of demand: PC BOM share hits 35-38%, shipments plunge — tengyanAI · 2026-11-10
- Two engineers burned $500K in API usage in one month — $6M annualized — bingxu_ · 2026-11-10
- Flide: exact state forking speeds decision-model agents up to 9.5x — NationalUniversityofSingapore · 2026-11-10