Chinese semiconductor makers intend to triple production of artificial-intelligence chips next year, the Financial Times reported, citing people familiar with fabrication plants that supply Huawei and other domestic clients. The expansion is designed to shore up supplies of advanced processors after US export controls sharply limited access to Nvidia’s most powerful GPUs. Higher output would back the ambitions of Beijing-based AI developer DeepSeek, which this week rolled out version 3.1 of its large language model. The update adds UE8M0 FP8 support, enabling the software to run efficiently on locally manufactured chips that require less compute and lower costs than imported alternatives. Industry executives said volumes of 7-nanometer-class AI accelerators could rise threefold under the plan, helped by state incentives and orders from cloud providers and handset makers. The push underscores China’s broader effort to build a self-reliant AI technology stack as tech tensions with Washington intensify.