384: Chapter 384 Space Computing Power: Nebula Online

In Hainan during April, the sunlight already carries the heat of early summer. In the coconut groves outside the Wenchang launch site, the chirping of cicadas rises and falls, sounding like a prelude to the impending rocket launch. On Launch Pad No. 3, the Tianlong-Medium Rocket stands quietly, the dew on its body glinting in the morning light. Unlike previous launches, the fairing is not printed with satellite patterns, but with a simple line of white characters—"Nebula"—beside which is a smaller line: "Space Computing Power Center · First Batch of 100 Units."

These are not satellites or space station modules, but servers. One hundred servers, packed into a specially sealed cabin, are about to be sent into an orbit of five hundred and fifty kilometers. This is the first time humanity has sent large-scale computing power into space.

Chu Tianming stands on the viewing platform, surrounded by the core team of Xingchen Aerospace. Today, he wears a light blue work uniform with the Xingchen logo embroidered on the chest, his hair slightly tousled by the sea breeze. At sixty, his face bears a new expression—not nervousness, not anticipation, but a nearly tranquil certainty.

“Today,” he says into the microphone, his voice broadcast globally, “we are sending one hundred servers into the sky. They will form humanity's first space computing power center in an orbit of five hundred and fifty kilometers.”

The large screen behind him lights up, and technical parameters appear one by one:

“Nebula” Space Computing Power Center

Deployment Orbit: 550 km (Sun-synchronous orbit)

First Batch: 100 servers

Individual Computing Power: Equivalent to 1,000 ground servers

Total Computing Power: Equivalent to 100,000 ground servers

Energy Consumption: 80% lower than ground-based (utilizing the low-temperature environment and efficient heat dissipation of space)

Communication Bandwidth: 10 Gbps (Laser communication)

Latency: Approximately 5 milliseconds to the ground

Reporters below record the data rapidly. A reporter from MIT Technology Review raises a hand: “CEO Chu, why is the computing power increased so much? A thousand times—this isn't something Moore's Law can explain.”

Chu Tianming smiles, a look of technical sincerity in his expression: “Space has no electromagnetic interference, no vibration, and no temperature fluctuations. Servers can run at full speed without frequency throttling. Add to that Xingchen's chip—the seven-nanometer optical quantum architecture—and it's that simple. On the ground, thirty percent of a data center's energy is spent on cooling. In space, the environment at minus one hundred and seventy degrees is a natural radiator.”

At 9:30 AM, the countdown begins. In the Command Center, technicians at hundreds of workstations hold their breath. On the large screen, data streams jump continuously—all systems normal.

“Ten, nine, eight, seven, six, five, four, three, two, one—Ignition!”

Flames erupt from the base of the rocket as the Tianlong-Medium slowly leaves the launch pad. Unlike the majesty of a heavy rocket, the medium rocket's ascent is more agile, like a silver arrow piercing the sky. Thirty seconds later, it passes through the clouds and vanishes from sight.

At 10:00 AM, the rocket and payload separate. On the large screen, one hundred points of light eject from the final stage of the rocket, slowly spreading out like a flock of birds just leaving the nest. They will automatically form a network over the next three days, connecting into a computing power array via laser communication links.

Chu Tianming watches those points of light and says softly, “From now on, humanity's calculations will be done in the heavens.”

Liu Peiqiang stands beside him. The white-haired veteran scientist adjusts his glasses: “One hundred units is just the beginning. A thousand next year, ten thousand the year after. By then, half of the ground-based data centers can be shut down.”

That same afternoon, in Mountain View, California. Google headquarters is nestled in the foothills of the Santa Cruz Mountains, with low-slung buildings hidden among trees like a closed-off technological kingdom. In CEO Sundar Pichai's office, the atmosphere is as tense as a war room.

On the large screen, the recording of Xingchen's press conference is playing. When Chu Tianming says, "the computing power of a single unit is equivalent to a thousand ground servers," Pichai's finger taps lightly on the desk.

“We need this.” His voice is calm, but his tone is beyond doubt.

The Chief Technology Officer frowns: “With the U.S. sanctions, we can't cooperate directly. The Department of Commerce will block any contract involving Xingchen technology.”

Pichai falls silent for three seconds and then says, “Then use the European subsidiary. Rent it in the name of Google Ireland. Ireland is an EU member and is not within the scope of the sanctions.”

The Chief Legal Officer shakes his head: “We can't bypass it. Xingchen won't accept it. And if we're discovered, the Congressional hearings will be more than we can handle.”

Pichai closes his eyes, and when he opens them again, there is a weary determination in his gaze: “Then lobby the government. Tell them that if American companies can't use space computing power, the Europeans will. By then, we'll have lost the AI race.”

That same afternoon, in Shenzhen. On the thirty-eighth floor of Xingchen Headquarters, three phones on Lin Wan's desk ring in turn.

The first is a formal inquiry from Google Ireland. Lin Wan glances at it and sets it aside.

The second is from Microsoft UK. She places it on top of Google's.

The third is from Amazon Luxembourg. She puts it on the very top.

Then she picks up the phone and dials Su Chen's number: “Chief Su, the three American giants have all come. They're using European subsidiaries to bypass the sanctions.”

Su Chen's voice comes through the receiver, as calm as if discussing what to have for dinner: “Tell them it won't work. Xingchen's customer audit looks at the ultimate beneficiary. The ultimate beneficiary of Google Ireland is still Google USA.”

Lin Wan smiles: “I thought so too.”

Su Chen continues: “Let them lobby their government. If the sanctions are lifted, they are welcome. If not, we aren't lacking for customers.”

He truly doesn't lack customers. The next morning, orders from Europe fly in like snowflakes.

The Max Planck Institute in Germany is the first to send an email: “We need Nebula's computing power for climate model simulations. One thousand hours; name your price.”

The French National Centre for Scientific Research follows immediately: “Gene sequencing data, urgent processing required. Five hundred hours.”

The email from ASML in the Netherlands is even more direct: “Lithography machine simulation requires massive computing power. Two thousand hours. Please provide a quote.”

Then comes Saudi Aramco: “Oil exploration data processing. Three thousand hours.”

National University of Singapore: “AI training. Five hundred hours.”

Within a day, research institutions and enterprises from Europe, the Middle East, and Southeast Asia have snapped up eighty percent of the computing power share. Lin Wan's assistant looks at the order list and can't help but say, “Director Lin, should we raise the prices?”

Lin Wan shakes her head: “No. Let the customers wait in line. The more people waiting, the greater our value becomes.”

April 3rd, New York. In an office building in Lower Manhattan, the offices of the high-frequency trading firm "Quantum Capital" are brightly lit. CTO Mark Zuckerman stands before a row of servers, data refreshing on the small screen in front of him at millisecond speeds.

His company specializes in cross-market arbitrage—making money from the tiny price differences between the four exchanges in New York, Chicago, London, and Tokyo. Traditional arbitrage relies on ground-based fiber optics. From New York to Chicago, the latency is about seven milliseconds. From New York to London, it's about forty milliseconds. Speed is money.

But today, he is testing something different.

He has built a system that sends part of the computing tasks via ground base stations to the Nebula Space Computing Power Center—on an orbit five hundred and fifty kilometers high, where one hundred servers are running at full speed in a vacuum at minus one hundred and seventy degrees. The calculation results are sent back to New York via laser communication.

Latency: 5.3 milliseconds.

Thirty milliseconds faster than ground-based.

Mark's eyes light up. He says to the CEO, “Boss, we're going to strike it rich.”

The next day, the team simulated one million transactions using historical data. The results were breathtaking—the ground system averaged a profit of 0.02% per transaction. The space computing system averaged 0.07%. A difference of only 0.05%, but multiplied by millions of daily transactions, it becomes an astronomical figure.

The CEO asks, “Is it compliant?”

Mark shrugs: “The rules don't say we can't use space computing power.”

April 5th, Washington. In the building of the U.S. Securities and Exchange Commission, an emergency meeting is underway. Chairman Gary Gensler looks grim; on the screen before him is a detailed report on Quantum Capital's test system.

“Several Wall Street firms are using Xingchen's space computing power for high-frequency trading,” his voice is low. “The latency is thirty milliseconds lower than ground-based. This gives them an unfair advantage. This isn't technical competition; it's a loophole in the rules.”

A commissioner asks, “Can we ban it?”

Gensler flips through the legal texts in front of him: “There are no clauses in the existing rules prohibiting the use of space computing power. But we can make an emergency amendment. Rule-making doesn't require Congressional approval; the Commission has the authority.”

That afternoon, the SEC issued an emergency notice: “Effective immediately, all U.S. financial institutions are prohibited from using overseas space computing facilities for securities trading. Violators will face heavy fines.”

In Quantum Capital's office, Mark looks at the notice and gives a bitter smile. He calculates that from testing to prohibition, it took only forty-eight hours. The thirty-millisecond advantage is gone.

IV. The Future of Computing Power

April 6th, Shenzhen. Su Chen sits in his office, looking at the SEC notice, and smiles.

Lin Wan stands to the side: “Chief Su, they've banned it. The path for high-frequency trading is cut off.”

Su Chen shakes his head: “It's not our path that's cut off, but theirs. Nebula's computing power was originally meant for scientific calculation, AI training, and gene sequencing. High-frequency trading was just an accidentally discovered application scenario.”

He pauses and picks up his teacup: “Besides, what's banned is financial trading. European scientific computing, Middle Eastern oil exploration, Southeast Asian AI training—they can't touch those. Those customers will be enough to sustain us for ten years.”

Lin Wan thinks for a moment and asks again, “What about the American companies? Will they lobby the government to lift the sanctions?”

Su Chen sets down his teacup and looks out the window. On the waters of Shenzhen Bay, a cargo ship passes slowly.

“Sooner or later. When European AI models run faster than theirs, when European gene sequencing is more accurate than theirs, when European climate simulations are more precise—Wall Street will lobby, Silicon Valley will lobby, even The Pentagon will lobby. Sanctions cannot stop technological progress.”

Outside the window, the sunset is falling. The sky over Shenzhen Bay is dyed an orange-red, and the distant Hong Kong skyline is faintly visible in the twilight.

Su Chen says softly, “Computing power is the oil of the future. Whoever controls computing power controls the future. And our computing power is in the heavens. They can't reach it, and they can't ban it.”

Lin Wan gazes at his back and says nothing.

In the deepest reaches of the sky, on an orbit of five hundred and fifty kilometers, one hundred servers are networking. Laser communication links flash between them like a silver spiderweb, operating silently in the vacuum. Every millisecond, hundreds of millions of calculations are completed within them—climate models, gene sequences, AI training, oil exploration. Humanity's computing power has left the ground for the first time.

And this web will only grow larger and larger.