This move was unprecedented.
China Star had officially deployed a supercomputer—Huaxing-1—as its primary cloud server.
That meant using a 2,000-teraflop (2 petaflop) supercomputer not for research institutions or the military but for commercial operations.
According to the current global rankings, Huaxing-1 could easily break into the top five supercomputers in the world.
And now?
China Star wasn't just using it.
They were using it for cloud computing. For consumer and enterprise applications.
No one—not in China, not globally—had ever done this.
Supercomputers were usually strategic assets, tightly controlled and funded by governments. Using one as a company's back-end server?
That would make national defense researchers lose their minds.
And yet, Haifeng looked over the humming, ice-cooled towers of Huaxing-1 and smiled in satisfaction.
"With this," he said, "China Star's R&D and server backbone will leap lightyears ahead of the competition."
China Star is conducting advanced artificial intelligence research led by AI expert Yan Xiping. The project required massive-scale data computation—and the only way to support it was a high-performance compute cluster.
At the same time, Haifeng was laying the groundwork for China Star's expansion into the internet industry.
Search engines. E-commerce. Streaming. Social media.
All of that required unshakable server infrastructure—and now they had it.
📍 Late November – China Star HQ
In his office, Haifeng was reviewing a competitor's new smartphone. He sighed.
"Xiaomi again…"
They had just released the third generation of their mainline phone series—and priced it, as always, at ¥1,999 (≈ $275 in 2025).
Zhang Yu walked in with a frown.
"This time, they didn't go with Qualcomm. They used a processor from Nvidia."
"That's… surprising."
Haifeng nodded thoughtfully.
Nvidia was well-known for its graphics cards—not phone chips.
Yet, Xiaomi boldly launched a mobile version of the Mi 3 powered by Nvidia's newest SoC.
"Qualcomm dominates the mobile processor space," Haifeng said.
"But their prices are sky-high—and they have some… unofficial rules.
Xiaomi must've been trying to escape their grip."
Zhang Yu frowned.
"You think it'll work?"
"Performance-wise, it scored decently. Not far off from Qualcomm's flagship chips."
"But the optimization? Heat? Power management?"
"Let's just say… we'll see soon enough."
Sure enough, by early December, news broke:
The Mi 3 was a disaster.
Despite strong specs, the Nvidia-powered Mi 3 suffered from:
Serious overheating
Lag and stutters
System instability
The tech forums roasted it. Even hardcore Xiaomi fans admitted they were disappointed.
Production slowed to a crawl. For the first time, Xiaomi phones were set on shelves unsold for five days straight.
"Usually, they sell out in 24 hours," Zhang Yu muttered.
"Now they can't even move 50,000 units."
It was a huge blow.
Even Xiaomi's founder, Lei Jun, called emergency meetings to consider halting production.
In the boardroom, Xiaomi's Vice President Lin Wen looked grim.
"Mi fans are furious. If this keeps up, it could damage our brand long-term."
He looked at Lei Jun.
"We still have the Unicom version of the Mi 3 with a Qualcomm chip. Should we shift focus back to that?"
Lei Jun was silent for a long moment.
"We tried to cut the cord with Qualcomm… but weren't ready."
"Maybe it's time to call them again."
Lin Wen hesitated.
"What about MediaTek? Their chip powers our Mi Hong model—it's not the best, but it's cheap, and sales were solid."
"Not as powerful as Qualcomm, but way better optimized than Nvidia."
Xiaomi was stuck.
They had tried to escape Qualcomm's grip by backing a newcomer. It was bold—but ultimately, a misstep.
Haifeng, reviewing all this, couldn't help but smile.
This is the difference between specs and systems.
Performance means nothing without control.