Live Stream 1980: Netizens Teach Me How to Make a Rocket by Hand

Situ Yuan's expression changed.

That wasn't the caution of an ordinary engineer when faced with a difficult problem, but rather the fervor of discovering a new continent.

He put the red pen back onto the drawing.

"Abandon the von Neumann architecture."

Zhang Bingqian's fingers paused.

Zhao Sihai remained silent.

Situ Yuan drew a dense grid on the drawing board.

"We use thick-film circuits to construct a non-Von der Rohe resistor cross array."

"There's no need to move data around."

"Each node resistor on this thick film board is both a memory and an arithmetic unit."

"The resistance value determines the weight of the decision logic!"

As he said this, he made a heavy dot in the center of the grid with a red pen.

"In other words, we directly burned the normal state, fault state, priority, and judgment logic of 100,000 nodes into this thick film resistor network."

"It's not written into memory."

"It's hard-coded into the physical structure."

Zhao Sihai's Adam's apple bobbed.

That's a really harsh statement.

It was so ruthless that it didn't feel like they were repairing a circuit board; it felt more like they were building a new nervous system for the spaceship.

Situ Yuan continued speaking, his pace quickening.

"As soon as the electrical signal from the external sensor comes in, the matrix multiplication is automatically completed the instant the current passes through this resistor array, using Kirchhoff's laws."

"The resulting voltage distribution is the final diagnostic report!"

Situ Yuan stared at Lin Xi, speaking slowly and deliberately:

"President Lin."

"This is no longer an ordinary maintenance tool."

"Given a large enough scale, this thick membrane becomes the physical neural network of the spacecraft."

The moment he finished speaking, you could hear a pin drop in the lab.

Only the teacup beside Zhang Bingqian emitted wisps of steam.

The next second, Zhao Sihai slammed his hand on the lab table.

"Snapped!"

The tweezers on the table even jolted.

"This architecture... is fucking stable!"

His face flushed with excitement, and he waved his rough, calloused hands:

"Silicon wafers are susceptible to particle impacts, but my thick ceramic film isn't!"

"Thick film resistors are made by burning metal paste onto ceramic at high temperatures."

"Don't even mention high-energy particles; even if you smash it with a hammer, as long as the ceramic doesn't break, its physical resistance won't change!"

"Logical links are directly made into physical links."

Zhang Bingqian understood and replied...

"It doesn't need to move data, so there's no transmission delay."

"The result is obtained the instant the current flows."

"This is true parallel processing."

As he said this, he looked up at Situ Yuan.

"The response speed may be hundreds or even thousands of times faster than that of a silicon-based CPU."

Zhao Sihai grinned, as if he had heard some incredibly good news.

"Isn't that just like firing your brain into ceramic?"

This statement is crude, yet shockingly accurate.

Three top technical talents from three generations—old, middle-aged, and young—combined their ideas in a cross-generational brainstorming session.

Lin Xi stood to the side, feeling a faint tingling sensation in her back.

Originally, he just wanted to build a doctor for the spaceship that could find the root cause of its problems.

But now, this plan has been overturned on the spot by three people from Red Star Semiconductor, and they have been put back together.

It is not a monitoring bus.

It is the spacecraft's neural network.

It's not about adding a patch to an old system.

It is about replacing the underlying architecture of future aerospace electronics.

One person designs the architecture, another considers the manufacturing process, and the third focuses on reliability.

On this drawing, the three lines truly interlocked for the first time.

In Lin Xi's mind, the live stream chat was in complete chaos.

Holy crap! Holy crap! Holy crap! I'm getting goosebumps!

[Physical hard coding + in-memory computing + ceramic thick film for radiation protection! What an amazing combination!]

This isn't just a troubleshooting bus; it's a prototype of hardware-level artificial intelligence!

[A graduate student from the University of Electronic Science and Technology of China is in tears. Why is it so difficult to achieve storage-computing integration in the future? Because the ecosystem, instruction set, and business model are already locked in; starting from scratch is too difficult.]

But in 1985, they were faced with a blank slate. Without the constraints of Western standards, they were starting from the ground up, forging a new path!

[In a parallel universe, China's aerospace industry has surpassed its predecessors in terms of hardware infrastructure! This is a resounding victory!]

Lin Xi took a deep breath and suppressed her turbulent emotions.

He walked up to Situ Yuan, placed his hands on the edge of the control panel, and his gaze was sincere yet sharp:

"Chief Engineer Situ."

"How long would it take to produce a usable engineering prototype of this thick-film neural network array?"

Situ Yuan pushed up his glasses and quickly went through the workload in his mind.

"The design of the blueprints took half a month."

"Old Zhao's sample sintering takes a week."

"Add compatibility testing to the underlying code..."

He gave a time,

"I'll get it done for you in two months at most."

Zhao Sihai shouted at the top of his lungs from the side:

"President Lin, don't worry!"

"Once Situ draws it out, I, Old Zhao, will not eat or sleep, but I will definitely print this neural network for you clearly!"

Zhang Bingqian picked up the teacup and gently tapped the rim with his finger.

"I'm responsible for keeping an eye on the logic gates and interface parts."

"If this thing is going to be a super high-tech product, we can't just pursue novelty; it must be stable."

"it is good."

Lin Xi nodded, rolled up the sketch again, and stuffed it into her briefcase.

He looked at the three key technical personnel in front of him and bowed deeply to them:

"This is not just a chip."

"It will be the most reliable brain in the future, keeping our eyes open hundreds of kilometers in the air."

Outside the window, the ice on the Haihe River in Tianjin is melting little by little.

In this somewhat crowded laboratory, China's aerospace industry, based on its electronic hardware foundation, quietly crossed the barrier of forty years and embarked on a peak path that future generations of aerospace professionals could only look up to.

......

Early April, Silicon Valley, Microsoft headquarters.

Inside the testing room, the rain pattered against the glass windows, the sound dense and cold.

The room was not brightly lit; only a few monitors emitted a pale blue light, making everyone's faces appear bluish.

William stared at the heavy monitor on the table, his brow furrowed.

Engineer David sat in front of the keyboard, pressed Enter, and dragged the mouse.

Several windows spread out on the screen.

There are no overlapping layers, no shadows, only harsh blocks of color dividing the screen into several squares.

When the mouse pointer moves, it leaves a noticeable trail.

"In order to circumvent the window overlap display patent of Red Star Technology's Taiji OS,"

David didn't turn around; he kept his eyes glued to the screen, reporting to his supervisor.

"We can only use a flat layout."

"Moreover, the underlying kernel still calls the single-task kernel of DOS."

He paused for a moment, then finally told the truth.

"There's just not enough time; we simply don't have enough time to rewrite it."

"Tai Chi OS... is too advanced."

The testing room was silent for a few seconds.

No one responded.

William could certainly tell.

The system on the screen is still a step away from a true graphical operating system.

No, the difference might be more than just a breath.

It's like giving the old DOS an ill-fitting suit; it looks decent from afar, but up close it's full of loose threads.

But the market won't wait for Microsoft to slowly regain its dignity.

William flipped through the report, becoming increasingly annoyed, and finally threw it on the table.

"I know this thing isn't perfect, but the market can't wait."

If Taiji OS can't produce a decent graphical operating system this month, it will attract all the software developers in the world.

By then, Microsoft won't even be able to get onto the poker table.

William looked up and asked:

"Have you installed the sample chips that Intel sent?"

"It's installed."

David answered immediately.

"80386".

"The computing power is terrifying."

"That's enough."

William reached out and patted the monitor casing, as if to encourage himself.

"The system's underlying lag was simply overshadowed by this new beast's computing power."

David opened his mouth, but in the end, he didn't say anything.