July 31, 2026 – DeepSeek is reportedly preparing a major expansion of its AI computing infrastructure by developing a large-scale data center in Inner Mongolia, signaling the Chinese startup’s growing investment in high-performance AI capacity.
According to a Bloomberg report citing people familiar with the matter, the company intends to deploy approximately 1 gigawatt (GW) of computing power in Ulanqab, a city in China’s Inner Mongolia Autonomous Region that has become an important hub for large-scale data center projects.
Sources who requested anonymity said the initiative is not limited to self-owned infrastructure. In addition to constructing its own facility, DeepSeek is also exploring the option of leasing additional computing resources from third-party providers to further expand its AI capabilities.
Two of the people familiar with the plans indicated that a portion of the new computing capacity is expected to come online by the end of next year or in early 2028, although the overall deployment timeline has not been fully disclosed.
It remains unclear which processors will power the planned data center. Nvidia continues to dominate the global AI data center market with its accelerator GPUs, while Huawei has emerged as one of China’s leading suppliers of AI infrastructure and computing hardware, making both companies potential reference points for large-scale AI deployments in the country.
The reported project highlights DeepSeek’s broader push to secure long-term computing resources as demand for AI model training and inference continues to rise, reflecting the increasingly fierce competition among Chinese AI developers to build and access large-scale computing infrastructure.
