China AI speaker says 2026 compute spend may stay under RMB 20B
teortaxesTex · x · 2026-07-23
A quoted talk says China will not try to compete with the US at the largest model scale yet, and instead focus first on model sizes it can afford to train and run.
Key points from the excerpt:
- The speaker expects no more than 20B RMB spent on compute in 2026.
- He says the largest current model activates about 800B parameters.
- Training that scale would require roughly 50K GB300s or 200K Huawei 950s.
- The quote argues you can brute-force a model that large, but not do enough research at that scale beforehand.
- He also says the remaining gaps in foundation-model competition are mostly cost, time, and user experience.
Related event: Chinese AI Chips Face Production and Compute Bottlenecks(4 posts)→
More from Infra
- Agentic RL setup hosts 135,000 task images and 1,000 concurrent rollouts — willccbb · 2026-07-23
- An AI engineer’s learning list covers caching, routing, RAG, and quantization — EmployerNegative5653 · 2026-07-23
- Huawei GPUs may only have a 3-year life, while NVIDIA chips keep a 5-year depreciation window — teortaxesTex · 2026-07-23
- Intel and AMD face rising China server CPU prices as lead times stretch to six months — firstadopter · 2026-07-23
- Kimi V4 is headed for native multimodality, but Moonshot says compute is the limit — teortaxesTex · 2026-07-23
- Huawei 950 supernode is described as a GB200/GB300 substitute, but capacity is still tight — teortaxesTex · 2026-07-23