> What infrastructure will these open weight models be trained on?

One, the infrastructure is being built for inference. Not training. If all we were doing was training on datacentres, I think America probably has enough already for near-term commercial needs.

Meituan’s 1.6T LongCat was trained entirely on Huawei training cards.

DeepSeek, GLM, Qwen and others are also actively working on similar replacement.

Chinese infrastructure, presumably.