ZH reported, citing a May 13 report from China Daily.
A new kind of infrastructure is emerging in China—one that no longer treats electricity and computing as separate systems.
Instead, they are being fused into a single architecture.
In this system, renewable energy is no longer just powering cities and factories. It is being directly engineered to fuel data centers, AI training clusters, and national computing networks.
This is the rise of what can be called China’s “bits and watts” economy: a structural convergence of digital computation (“bits”) and physical energy (“watts”).
From Power Grid to Computing Grid
Traditionally, electricity flows through a layered system:
power generation → transmission grid → distribution network → end users
But China’s new model is beginning to bypass parts of this structure.
In a major new development in Ningxia, a large-scale photovoltaic power station has been directly connected to computing infrastructure through dedicated transmission lines.
Instead of feeding electricity into a general grid, renewable energy is now being routed straight into data center loads.
This represents a subtle but fundamental shift:
electricity is no longer just distributed — it is being assigned to computation.
The East–West Computing Architecture
At the core of this transformation is a national strategy known as the East Data West Computing initiative.
The logic is spatial and industrial at the same time:
- Western China: abundant land, solar, wind, and low-cost energy
- Eastern China: dense population, AI companies, cloud computing demand
Instead of forcing data centers to cluster near cities, China is effectively relocating computation to where energy is cheapest and cleanest.
This is not just an infrastructure project.
It is a geographic redesign of digital capitalism.
The Emergence of Direct Green Power Supply
One of the most significant innovations in this system is “direct green power supply.”
In the traditional model, renewable energy enters the public grid and becomes indistinguishable from fossil-fuel electricity.
In the new model, energy flows are becoming more traceable and isolated:
- dedicated transmission lines connect renewable plants directly to data centers
- “point-to-point” supply bypasses intermediate grid mixing
- energy storage systems smooth fluctuations between solar and wind cycles
This creates a highly controlled energy environment for computation.
The result is a measurable increase in the “green power ratio” of digital infrastructure.
Why Data Centers Are Becoming Energy Infrastructure Nodes
Data centers are no longer just computing facilities.
They are becoming hybrid energy–computing nodes.
Their role now includes:
- consuming massive, stable electricity loads
- anchoring renewable energy investments
- enabling predictable demand for wind and solar expansion
- optimizing carbon intensity of digital services
In this model, computation is not just powered by energy—it structurally organizes energy demand itself.
This reverses the traditional hierarchy.
Energy no longer simply enables computing.
Computing now shapes energy deployment.
Solving the Geography of Energy and Data
One of the key constraints in digital economies is geographic mismatch:
- renewable energy is often produced in remote regions
- data centers are concentrated in economically dense coastal cities
The Chinese model directly addresses this by relocating computation closer to energy sources.
This reduces:
- transmission losses
- electricity costs
- carbon intensity
- grid congestion
More importantly, it creates a synchronized system where energy production and digital consumption are co-located across regions.
Stability Through Energy Complementarity
A defining feature of the Ningxia project is its hybrid energy design:
- solar power dominates during daylight hours
- wind power takes over at night
- energy storage smooths fluctuations
This creates a continuous, 24-hour supply loop tailored specifically for computing workloads.
It reflects a shift from intermittent renewable deployment to system-level energy engineering.
Instead of treating renewables as variable inputs, they are being orchestrated into stable industrial supply chains.
The Strategic Logic: Energy as Digital Infrastructure
The broader implication is that energy is being redefined as digital infrastructure.
In this model:
- electricity is not a commodity
- it is a compute input
- and increasingly, a strategic resource for AI systems
As China expands its national computing capacity, energy planning and digital planning are becoming inseparable.
Policy and infrastructure are converging around a single idea:
computing capacity is constrained not only by chips and algorithms, but by watts.
The Cost and Carbon Dimension of AI
AI systems are becoming energy-intensive at scale.
Training large models, operating cloud systems, and running inference networks all require continuous, stable power.
By linking renewable energy directly to computing infrastructure, China is attempting to solve three problems simultaneously:
- energy cost volatility
- carbon emissions from computing
- regional imbalance in infrastructure deployment
This creates a structurally optimized system for AI-scale development.
From Industrial Power to Cognitive Power
Historically, industrial power was defined by steel, manufacturing, and physical production capacity.
The emerging paradigm is different.
Power is now defined by:
- computing capacity
- energy efficiency
- and the integration of digital and physical systems
In this sense, “bits and watts” are becoming the dual currency of industrial competitiveness.
Conclusion: A New Infrastructure Logic for the AI Era
China’s integration of renewable energy and computing infrastructure represents more than an energy transition.
It represents a structural redefinition of how digital economies are powered.
In the emerging “bits and watts” economy:
- energy is no longer downstream
- computing is no longer independent
- infrastructure becomes a unified system
The result is a tightly coupled architecture where green power is not just supporting AI—it is becoming its foundational fuel.
As this model expands, it may redefine not only how AI systems are built, but where and why they are built at all.