Moore Threads Accelerates Three "AI Factories" to Face the Token Economy Era

Moore Threads, the Chinese GPU manufacturer, announced today that its three "AI factories" are accelerating toward deployment. According to reports via Sina, the strategic initiative is designed to meet the demands of the emerging token economy era by providing domestic AI computing infrastructure for large-scale inference and training workloads.

The "AI factory" concept, as proposed by Moore Threads, refers to standardizing GPU computing resources into turnkey AI infrastructure services, offering enterprises and developers an integrated solution spanning model training to inference deployment. The approach bears conceptual similarities to NVIDIA's DGX SuperPOD/AI factory model, but is built on domestic GPU technology.

In the token economy era, every inference call to an AI model generates token consumption, with computing demand shifting steadily from training to inference. Moore Threads'布局 is essentially a strategic play to capture position in the inference computing market.

The company has previously launched multiple consumer and data center GPU products, establishing a significant position in China's domestic GPU sector. The accelerated rollout of three AI factories marks an acceleration of the company's strategic transformation from a pure chip design firm to an AI infrastructure operator.

For China's AI industry, the development of domestic AI computing infrastructure holds particular significance. In the current international environment, a self-controlled computing base has become critical to sustaining China's AI industry growth. If Moore Threads' AI factories achieve commercial-scale operation, they could provide a vital alternative computing resource for domestic AI companies.

Key factors to watch include the specific locations and capacity of these three AI factories, and whether leading AI companies will be among the first tenants.