Who would have thought that Ulanqab would also benefit from the AI dividend.
Recently, I saw Goldman Sachs, Bloomberg, and Wired discussing China's AI, and surprisingly, they all mentioned Ulanqab.
They say this small town in Inner Mongolia is becoming a major hub for computing power in China's AI infrastructure landscape.
Ultimately, the competition in AI boils down to competition in energy.
You see Musk has been talking about space data centers all year, one very important reason being that the electricity required by future AI will be enormous, and by moving to space, it can directly utilize a large amount of solar energy.
OpenAI is the same, with Sam Altman directly stating that the essence of their company is to convert electricity into useful intelligence.
Why Ulanqab then?
Ulanqab meets several conditions: 1) The weather is cold. 2) Electricity is cheap. 3) There is an abundance of green energy. 4) It is close to Beijing.
I'm quite familiar with Ulanqab, and I always make it a point to visit Huanghuagou during the summer holidays, so I must write about it in detail.
To put it this way, Ulanqab has basically two seasons all year round, one is winter and the other is summer.
And the winters are extremely long. I checked the data — the average annual temperature is only 4.3°C.
I recall a year when I went to the Ulanqab grassland in mid-September and was frozen stiff at night, shivering with cold.
I asked a friend, and he said that in Ulanqab, heating preparations usually start soon after the National Day holiday, and continue until mid-April the following year, lasting around half a year.
The weather is truly hot for only two months, July and August.
Ulanqab is at an elevation of approximately 1,400 meters and is situated on a pathway for cold air from Siberia to move south, with cold air from the north passing through the Mongolian Plateau, resulting in not only low temperatures but also strong winds.
The cold and windy conditions may not be comfortable for the local residents living in Ulanqab, but they are actually advantageous for data centers. The cold weather allows for the utilization of natural cooling sources for most of the year, reducing heat dissipation costs.
The strong winds also indicate that this area has excellent wind power conditions.
A cool breeze that not only helps data centers to dissipate heat but also generates electricity at the same time.
And the electricity is also cheaper here.
Inner Mongolia is originally an energy-rich province, with abundant wind power, photovoltaic power, and sufficient coal-fired power.
Places like Ulanqab have seen their combined wind and photovoltaic power generation approach nearly half of the city's total power generation by 2025, with wind and photovoltaic power continuing to increase in recent years.
When it comes to AI data centers of this scale, just how important is the cost of electricity? In our daily lives, we might think that a difference of a few cents per kilowatt-hour is negligible. However, for GW-level data centers, this is not the case at all.
Let me do a rough calculation for everyone.
Assuming an industrial park has an average load of 1GW, with an average load factor of 90%, it would consume approximately 7.884 billion kWh of electricity per year.
In August 2026, the agency purchase price of electricity for Mengxi Grid was approximately 0.3234 yuan per kilowatt-hour, while that of Jibei was around 0.3857 yuan per kilowatt-hour, with a difference of just over 6 cents.
But it's just over 6 cents. 78.84 billion kilowatt-hours translates to around 491 million yuan per year. In other words, when you reach the GW level, saving a few cents per kilowatt-hour can mean hundreds of millions of yuan per year.
Additionally, as mentioned earlier, Ulanqab has a significant advantage in that it has an abundance of green electricity.
Green electricity, in simple terms, refers to electricity generated from renewable energy sources such as wind power and photovoltaics. Why is this becoming increasingly important for AI data centers? It's the same calculation: a 1GW park consumes nearly 8 billion degrees of electricity per year.
If these electricity sources rely heavily on coal-fired power, the resulting carbon emissions will be extremely high.
Major global companies are now focusing on carbon emissions, and when Google, Microsoft, and other companies build data centers, they are also very concerned about the proportion of electricity coming from clean energy sources such as wind power and photovoltaics.
It can even be predicted that in the future, when discussing the cost of a model, in addition to the cost of tokens, people may also start to pay attention to the energy consumption and carbon emissions behind these tokens.
Some may ask, what about places like Ningxia Zhongwei and Xinjiang, where electricity is also very cheap, and the wind and solar resources may be even more abundant than in Ulanqab.
This highlights another advantage of Ulanqab. For those who frequently take the train, they may have heard Ulanqab's city slogan: "One step west of Beijing is Ulanqab."
It's only about 350 kilometers from Beijing, and can be reached in roughly two hours by high-speed rail.
Beijing is one of the most densely populated areas for domestic AI companies, research institutions, and internet companies. If you continue to move computing power to Ningxia and Xinjiang, electricity may be even cheaper and land more abundant, but network latency will also increase.
When working on cloud computing in the past, the western data center mainly undertook storage backup, offline computing, and other non-urgent tasks, and a delay of tens of milliseconds had no impact.
The training of large models is similar, once the training task is initiated, it takes several days or weeks, and is not sensitive to the latency from Beijing to the data center.
However, AI is now using more and more computing power for inference. When an agent performs a task, it constantly needs to call models and various services in between. These calculations that directly face users are extremely sensitive to latency.
This is also Ulanqab's most significant advantage compared to Ningxia and Xinjiang, as electricity is already cheap and it is only a few hundred kilometers away from Beijing.
According to the data, Ulanqab started building a direct optical cable to Beijing in 2017, and later constructed a second one.
According to public data, the network latency from Ulanqab to Beijing can be as low as 5 milliseconds, with one route taking approximately 4.2 milliseconds.
Ulanqab's advantages are thus perfectly suited.
Heading west, electricity in Ningxia and Xinjiang may be even cheaper, but it would be farther away from Beijing. Heading east, the distance to Beijing would be shorter, but the costs of electricity and land would be higher.
I just excitedly told my friend in Ulanqab about these things, and he only asked me one question:
This will bring how many job opportunities to Ulanqab?
To be honest, the employment opportunities brought by the computing power center are extremely limited.