Sources: DeepSeek CEO Liang Wenfeng says training on Huawei chips is one of DeepSeek's biggest bets and Huawei is set to deliver training chips in Q4 or Q1 2027 (The Information)

DeepSeek’s push to train AI models on Huawei chips signals a broader shift toward domestic semiconductor supply chains, with potential implications for AI cost structures, vendor dependence, and geopolitical resilience. For CIOs and technology leaders, this suggests that enterprise AI roadmaps may increasingly need to account for local-chip ecosystems, model-training performance tradeoffs, and long-term diversification away from a small set of global GPU suppliers. IT organizations should expect more pressure to evaluate non-NVIDIA options, adapt infrastructure and software stacks for heterogeneous hardware, and align procurement strategies with regional supply and policy constraints.