I bought a prison in the United States.

"After you finish eating, come with me to the factory. Craig is already waiting for you there."

"Starting today, you will work eleven hours a day until the prototype of the monitoring system is completed."

"If you behave well, I will keep my promise and give you points and a reduced sentence certificate. If you don't behave well..."

Lingo paused, his gaze sweeping over the three faces.

"You know the consequences."

The three prisoners kept their heads down, silently eating their sandwiches without saying a word.

They all knew in their hearts that Ling Ge was not joking.

In this prison, he says what he means and he does what he says.

Lingo turned and walked out of the solitary confinement cell. Sunlight streamed in through the corridor windows and fell on him.

Hamon was standing at the end of the corridor, waiting for him.

"How is it?"

"I asked Hamon."

"It's done, they've given in."

Lingo said:

"Starting tomorrow, the technology research and development team will officially begin work."

Harmon stared at Ringo with suspicion; even now, he wasn't sure if his actions had any meaning.

"I still want to ask, do you really think those people are confident they can develop a completely new surveillance system within two months?"

Lin Ge chuckled softly:

You'll find out then.

The next morning at eight o'clock, the three members of the technology research and development team were summoned to Lin Ge's office again.

When Craig came in, he was holding a well-worn copy of "Introduction to Solid State Electronics," with several folded drawings tucked between the pages.

His habits, which he developed at Lockheed, were deeply ingrained; whenever he went to see the person in charge, he would bring relevant technical documents with him for easy reference.

Omar and Morris followed closely behind, each carrying a notebook filled with calculations written in pencil.

The three of them stood side by side in front of the desk, in the same position as when they were called in yesterday.

But the atmosphere has changed.

Today they were wearing relatively clean prison uniforms, their hair was combed, and Omar had even shaved his beard.

Ringo sat behind his desk, with Craig's technical assessment report spread out in front of him.

Harmon stood by the window, holding a coffee cup, adopting the posture of an observer.

"The reason I called you here today is to finalize the technical roadmap."

Lingo turned to the system architecture page of the report:

"Craig, have you estimated how long it would normally take to develop a complete closed-circuit monitoring system from scratch?"

"Eight months to a year, if the team is outside the prison and has normal equipment procurement channels and testing environment."

Craig answered without hesitation; in fact, this was already a rather conservative estimate.

The actual timeframe is often longer than that, since research and development involves an element of luck.

After hearing this, Lin Ge shook his head:

"No, your workplace is in this prison, and I need to see a working prototype by mid-December."

Omar and Morris exchanged a glance.

Neither of them spoke, but the eye contact itself conveyed a wealth of information.

They had already discussed the feasibility of this timetable privately yesterday.

The conclusion was that they believed it could work, but only if the warden was willing to provide some resources that they didn't expect to have.

"Last night I went through the entire system architecture again."

Upon hearing Ringo's request, Craig frowned, took a folded sheet of paper from his breast pocket, and spread it out on Ringo's desk:

"The main technical challenges at present are mainly in three areas."

"First of all, there's the image sensor in the camera. All the surveillance cameras available on the market now use a CCD with an image intensifier tube, which costs two to three thousand US dollars."

"If we buy ready-made cameras to install, the budget for the cameras alone would be close to 40,000 yuan for the entire prison with sixteen cameras."

"So we can't buy Honeywell's finished products?"

"Ling Ge said."

"right."

Craig nodded:

"If we assemble it ourselves and replace the CCD with a CMOS image sensor, we can reduce the cost to one-tenth."

"However, CMOS imaging quality is not as good as CCD in low light conditions, and corridor lighting is precisely a low-light environment, so some compensation needs to be made in the circuit design."

Lin Ge took a piece of paper out of the drawer.

He drew that last night in the motel; it was a very simplified cross-sectional diagram of a photosensitive unit on a piece of paper.

The pixel layer is on top, and the metal wiring layer is on the bottom.

Its most distinctive feature is that it allows light to enter the photodiode from the back, abandoning the traditional design where light passes through layers of wiring from the front.

He pushed the paper in front of the three people.

Craig glanced down, not initially paying much attention.

What's so special about a circuit diagram drawn by a prison boss?

But when he scanned that cross-sectional structure, his pupils suddenly contracted, as if he had been electrocuted.

"This is..."

His gaze lingered on the drawing for about five seconds, then he picked it up and looked at it more closely.

"What kind of structure is this?"

Craig's voice became very focused.

"Reverse the light-receiving surface of the photodiode."

Lin Ge tapped the pixel layer on the drawing with his fingertip:

"In traditional front-illuminated CMOS, light must first pass through the gaps in the metal wiring layer before reaching the photodiode."

"The wiring layer will block more than 50% of the incident light."

"If we reverse the process and allow light to enter the photodiode directly from the back of the silicon wafer, the amount of light entering the diode can be increased by at least 40%."

"The signal-to-noise ratio also increases under low light conditions."

Some of the technical ideas for this plan came from a memory that remained deep in Lingo's mind.

In his past life, when he was doing cross-border e-commerce, he had a long-term cooperation with a security equipment manufacturer in Shenzhen, specializing in exporting low-end surveillance cameras to Southeast Asia.

Although that collaboration did not last, the entire product line was halted after the chip supply was cut off.

But his experience of spending two whole summers in the factory with the engineers left an indelible impression on him regarding the basic principles of CMOS photosensitive structures.

Although he couldn't recall the specific process parameters, he at least remembered the idea of ​​"reversing" the process.

In the 21st century, back-illuminated CMOS sensors have become standard in smartphone cameras.

However, in 1984, no one had publicly proposed this idea.

For engineers, this is almost like a completely new worldview.

Craig handed the blueprints to Morris.

Morris took the blueprints with both hands, examining every detail from the perspective of technological feasibility.

He spent eleven years in IBM's hardware division, reviewing hundreds of design drawings and developing an almost intuitive sense of judgment.

"A back-illuminated structure?!"

After one glance, he wore an incredulous expression. He stared at it for a long time before slowly speaking:

"This makes perfect sense in principle, but there is a huge problem in its actual implementation."