JD Cloud to Build 100,000-GPU Computing Cluster with Moore Threads

JD Cloud has announced plans to build a significant computing cluster featuring 100,000 general-purpose GPUs from Chinese chipmaker Moore Threads. The project is intended to support the training and inference of large AI models, as well as embodied AI applications.

The initiative, unveiled at JD's 2026 Global Technology Explorers Conference, will make computing capacity available to companies across various industries. JD Cloud stated the project aims to establish large-scale computing infrastructure based on domestic hardware and facilitate the commercial deployment of physical AI.

The company highlighted this as the first deployment of a 100,000-GPU core computing cluster utilizing Chinese-made GPUs by a major domestic AI cloud provider.

This development is part of China's ongoing efforts to bolster its domestic semiconductor industry and AI capabilities, potentially reducing reliance on foreign technology and fostering local innovation.