B300 spot prices hit $2.2M per unit in China, 3x premium pushes domestic chips into the推理 sweet spot
aigclink · x · 2026-09-11
A Chinese operator reports B300 spot prices in China reaching 16M RMB per unit vs under 5M RMB overseas — a 3x premium.
Key takes:
- The premium triples per-token inference cost on B300 in China, pushing it to parity with domestic chips
- Domestic accelerators become the optimal choice for inference operators not because they improved, but because of Nvidia's distorted pricing — which will drive domestic chip shipments and iteration
- Author predicts the balance shifts to a price war after Dec 2026, with supply-demand normalizing by mid-2027
- Context: national cluster vacancy rates exceeded 50% before Oct last year; the author's own clusters ran 30% vacant
Bottom line: the good times last about a year.
More from Infra
- SemiAnalysis: OpenAI reportedly negotiating ~1GW TPU commitment while seeking sites — dylan522p · 2026-10-28
- I tested 11 video upscaling methods for AI video: local, Topaz and cloud compared — lopas99 · 2026-10-28
- Thermo-backed CTO says AI is in its dial-up phase, touts disaggregated compute for an ultrafast 'broadband moment' — beffjezos · 2026-10-28
- DeepSeek-V4.1-Flash Q2 runs at 41 tok/s on one RTX 3090, 90% token match vs API — musicurgy · 2026-10-28
- USearch v3 RC ships after 1,000 days, its libraries now embedded in 550K-star projects — viglovikov · 2026-10-28
- SK Hynix says memory demand will outpace supply through 2030 — Chemical_Hawk_6307 · 2026-10-28