Computing Power Takes Center Stage at 2026 Two Sessions, Shaping Digital Economy Future

Computing power has emerged as a high-frequency keyword among deputies and members during the 2026 national Two Sessions, with discussions focusing on its layout, construction and technological breakthroughs to underpin the in-depth integration of the digital economy and artificial intelligence.

The Government Work Report proposed implementing new infrastructure projects such as ultra-large-scale AI computing clusters and computing-power-electricity coordination, strengthening the monitoring and dispatching of the national integrated computing power network, and supporting the development of public cloud services.

As artificial intelligence evolves from generative AI to intelligent agents, demand for computing power in large model training and inference is growing rapidly. Scenarios including autonomous driving, industrial internet and telemedicine are also imposing increasingly urgent requirements for low-latency and high-reliability computing power.

Against the backdrop of advancing the “Eastern Data and Western Computing” project, accelerating the formation of computing power networks, pursuing technological independence and controllability, and promoting inclusive sharing of computing power, deputies and members at the Two Sessions put forward suggestions on key issues, jointly mapping out a blueprint for the high-quality development of China’s computing power industry.

Optimizing the national integrated layout and improving the “unified network” for computing power dispatching have become key priorities. A CPPCC member suggested accelerating the optimization of the national computing power strategic layout and building a national computing power system featuring integration, classification, grading and scenario-driven development.

He put forward specific measures from four aspects: first, constructing a three-level computing power architecture of “national-regional-edge” to implement precise functional layout; second, establishing classified guidance and unified dispatching mechanisms to promote the efficient allocation of computing power resources; third, strengthening key technological research and development and network fee reduction to consolidate the foundation for integrated operation; fourth, deepening computing-power-electricity coordination innovation and the application of market tools to promote the green and inclusive development of computing power.

An NPC deputy emphasized the need to accelerate the implementation of the national unified power market system, support AI computing centers in piloting cross-provincial and inter-regional electricity market transactions, and improve the efficiency and benefits of computing-power-electricity coordinated allocation.

In addition, he proposed supporting direct connections between computing power centers and power generation enterprises involving green electricity, hydropower and nuclear power, and providing policy support in terms of transmission and distribution fees and project approval to effectively reduce energy costs and boost the rapid development of China’s artificial intelligence industry.

It is evident that China’s computing power construction is shifting from “scale expansion” to “quality and efficiency improvement”, and from “single-point layout” to “whole-network coordination. Accelerating the construction of an efficient, inclusive, green, safe and independently controllable national integrated computing power network is well-timed.

Notably, there is currently a gap in dedicated clusters for inference tasks in China’s computing power centers, with room for optimization in the supply-demand adaptation between regions and an urgent need to break through dedicated inference chip technologies.

In response, a CPPCC member suggested that the state issue guiding policies for inference computing power layout, establish a “national overall planning + regional refinement” inference computing power layout system based on local scenario density, computing power gaps and energy security capabilities, and build low-latency and high-density inference computing power clusters in key industrial agglomeration areas.

The value of computing power lies in its application, which plays a crucial role in deepening inclusive application and empowering the digital transformation of various industries. With the implementation of policies and joint efforts of the industry, China’s computing power infrastructure will become more improved, core technologies will make faster breakthroughs, and application scenarios will continue to enrich, making it a “smart engine” driving high-quality development.