Chen Cheng was shocked.

One to forty-five; this ratio caught him off guard.

Li Na shrugged. "There's no other way. If we want the land, this is what we have to do."

However, the actual cost isn't that high. Calculating based on 50,000 mu, the initial restoration investment is roughly between 75 million and 250 million, depending mainly on the difficulty of the restoration.

After that, the annual maintenance cost is roughly between 3 million and 10 million.

Chen Cheng nodded, immediately thinking of "under-panel" and "inter-panel" planting, which no one was using yet but would become very popular in the future.

Photovoltaic panels need to be rinsed and cleaned with water regularly, and the water will fall directly under the panels.

Plus, after being shaded by the photovoltaic panels, evaporation under the panels is reduced.

Even if economic crops aren't planted, sheep can be raised to keep the weeds low, preventing them from blocking the photovoltaic panels and affecting power generation efficiency.

The only problem is that the photovoltaic field is estimated to only cover about 1% of the total area.

Without changing the evaporation rate, only this portion of the area can be truly restored.

The remaining 99% of the area will have to rely on maintenance funding to fight against nature.

Only when the soil and water are truly conserved, and the difference between evaporation and precipitation is changed, can the restoration be considered a true success.

Chen Cheng thought for a moment, then shrugged it off. It wasn't that much money per year.

Fighting against nature is endless fun, isn't it? It can be considered a bit of entertainment.

Chen Cheng thought for a moment. "The electricity is solved, but what about the water?"

For the data center, even if we use the most advanced liquid cooling technology...

The temperature of the coolant will still rise, and at that point, water will be needed to cool the coolant down.

And the amount of water required is astronomical.

The industrial land that North of the Great Wall can provide will likely have a major problem with water.

Li Na thought for a moment. "The data center experts mentioned a solution."

We can dig an artificial lake with a depth of over 10 meters and lay the coolant pipes at the bottom of the lake.

In the lakes of North of the Great Wall, the deep water layer above 8 meters stays at a stable temperature of about 7-8°C.

Using a 10-meter water layer for the cooling pipes means the temperature change caused by heat dissipation is basically negligible.

It also won't change the evaporation rate on the lake's surface.

Chen Cheng nodded. "What about the investment budget?"

I looked into the earthwork excavation and transportation. To meet our water needs, depending on the geological conditions, it would cost roughly between 30 million and 45 million.

Adding in the loss of rainwater due to evaporation, we might also need a pipeline to draw water from a water source.

The initial investment for the water problem should be controllable under 100 million.

Compared to our total investment, this portion is relatively small.

Chen Cheng pondered for a while. "Let's proceed in this direction. Get a few plans ready as soon as possible."

Half a month ago, I reached an agreement with Yinda.

Before October, Yinda will deliver 6,000 new flagship computing cores to ZEN.

Li Na thought for a moment. "6,000 units? The company doesn't have enough money in its accounts, right?"

Li Na was now very familiar with industry information, so as soon as Chen Cheng mentioned it, she knew the price.

Chen Cheng chuckled. "I will inject another 3 billion USD into ZEN."

Li Na's eyes widened. "How much?"

Zou Yu, sitting opposite them, found Li Na's reaction hilarious but didn't dare to make a sound.

Chen Cheng nodded. "I made a windfall recently."

The three or four-month gap is just enough for the training of ZEN 1.0.

After the new cards arrive and we finish assembling the new servers, we'll swap out the cards in the old servers.

Keep the old cards; don't sell them. We might need them later.

Li Na swallowed hard, unable to snap out of her shock for a long while.

Chen Cheng continued, "Before April next year, another 7,500 custom 8-card servers will be delivered to ZEN."

This time, Li Na was even more stunned. "We're going straight to 60,000 cards in the second phase?"

She knew about Chen Cheng's five-year data center plan.

But she hadn't expected the second phase to directly complete 60% of the future progress.

Chen Cheng chuckled. "Before December next year, another 5,000 custom servers will be delivered. I'll inject funds into ZEN for this batch next year."

Li Na was numb. "In such a rush?"

We have no choice but to be in a rush.

We need to maintain our first-mover advantage while stocking up on enough cards before the embargo.

If ZEN releases 1.0 and the US discovers that all parameters have been surpassed, then advanced process computing cards will definitely be banned.

He wasn't even thinking about the next generation of computing cards anymore.

By the time the next generation arrives, Altman will have released 3.5, causing a sensation in the industry, with everyone shouting that the AI era has arrived.

The US will immediately turn around and embargo both the new and old computing cards.

All ZEN can do now is develop stealthily while letting Company Name hoard cards with the image of a "foolish and rich" nouveau riche.

Chen Cheng bought the 7,500 custom servers in the name of Company Name training quantitative models.

What Yinda thinks doesn't matter; they only want to make money.

Besides, Chen Cheng is now a shareholder, so there's even less need to worry about extra complications.

Plus, in a month or two, when Company Name's 13F report is released, the terrifying growth in the fund's net value will make the outside world speculate that Company Name's quantitative model has found the answer to cracking the market.

Several seemingly unrelated events will be guided to connect together.

And then, hide the most correct answer in plain sight.

Before the embargo, Chen Cheng will continue to hoard through various channels in other forms.

How much can be hoarded through unofficial channels depends on fate.

When enough cards are hoarded, that will be the time to release ZEN 2.0 and deliver a fatal blow.

At the very least, a computing cluster of 100,000 cards will be enough for ZEN to use for ten years.

If domestic alternatives haven't emerged in ten years, then Chen Cheng might have to enter the semiconductor game himself.

Li Na digested the shock Chen Cheng had given her for a while.

She quickly sorted out the order, priority, and urgency of the various matters.

She nodded. "Okay, I'll make a schedule later."

Chen Cheng nodded. "Everything regarding ZEN is top secret. We need to raise the level of counter-espionage."

The model training hasn't started yet, so no one knows the answer. Once it starts, some of those who know the inside story might not be so well-behaved.

Whether it's bragging or intentional betrayal, I don't want to see any information leaks.

Li Na exhaled and nodded heavily. "I understand."

Chen Cheng smiled, reached out, and grabbed her hand. "Keep it up!"

It wasn't that Chen Cheng wanted to pressure her.

Since she sat in this position, she had to bear the pressure that came with it.

Among all of Chen Cheng's women, looking at it now, Li Na, as the Vice President of ZEN AI, is the one with the highest position and the most power.