You promised to build cars, so why did you go into the military industry?

Chapter 79 Others walk their dogs, Li Ran walks his car.

Having acquired a super factory without spending a single penny, Li Ran was in such a good mood that he couldn't hide it, and his steps were even lighter than usual.

After assigning two project managers to the construction site to oversee the progress, he began to plan how to use the future production capacity.

The super factory will be able to officially start production in 18 months, that is, in early 2017.

At that time, we can take advantage of the new production line to give the Lingyun SU7 a major makeover, and then launch a brand new model to make a good start.

The timing was perfect.

At 10:00 AM on May 15th, CNSA announced the first test date of the reusable version of the Long March 6.

The news immediately drew global attention, with everyone watching to see if CNSA would become the third organization, after United Industries and SpaceX, to master controlled rocket recovery technology.

After all, although United Industries got there first, it is ultimately a private enterprise, while the involvement of the state-owned enterprise represents a completely different level of power.

At 9:00 AM on May 20th, at the Qilianshan Satellite Launch Center.

The Long March 6-1 rocket stands quietly on the launch tower, sunlight shining on its body, giving it a uniform silvery-white sheen.

Unlike the elegant and slender shape of the Joint One, the upgraded Long March 6-1 has a diameter of 4.5 meters. The entire rocket looks robust and solid, and the sense of oppression is much stronger when standing under the launch tower and looking up at it.

Lin Yu, who participated in the entire process of rocket modification and debugging, was in a very calm state of mind at this moment.

In fact, he's quite experienced in rocket recovery, having successfully recovered the Joint One rocket so many times. He's seen it all.

It's just a reuse test, nothing to be nervous about.

As long as we can get valid data, even if the Rockets ultimately fail and collapse, it's all worth it.

That's a price that must be paid during the research and development process.

When it comes to rocket recovery, looking back and forth won't get you far.

With some time still before the launch window, Lin Yu got up and walked to his workstation, patting the shoulders of several teammates who had been working overtime with him: "Relax, don't feel pressured."

This test launch was originally intended to collect data; it would be best if it could be recovered intact, but it wouldn't be a loss if it couldn't be recovered.

"As long as I have the data to back me up, I guarantee I'll get it back completely intact the second time."

Hearing this, Lü Chaoping, who was signing documents nearby, paused for a moment, then shook his head with a smile: "The little lackey who used to follow me around asking all sorts of questions now speaks with the air of a master."

He pushed the signed documents aside, his tone tinged with emotion: "This modification has truly tapped into the full potential of the YF-100."

However, the unique gas generator flow channel layout of the ZK-308 you envisioned gave me some new ideas.

"Once the recovery mission is over, would you mind if I stayed at United Industries for a while?"

Lin Yu's eyes lit up: "Not at all! As long as you nod, our external researcher position is always open for you."

Lü Chaoping waved his hand: "That won't do, there are rules."

Lin Yu immediately realized what was going on, nodded, and didn't say anything more. After all, Lü Chaoping was the chief designer of the YF-100, and his identity was sensitive. Some restrictions were hard to overcome.

An hour later, a breaking news report was broadcast: "—The Long March 3 reusable carrier rocket successfully launched at 9:35 a.m. today from the Qilianshan Satellite Launch Center, successfully sending three experimental satellites into their predetermined orbits of 300 to 500 kilometers. However, during the recovery and landing phase, the rocket's attitude became abnormal, deviating from the planned landing site and crashing. There were no casualties."

The rocket debris and black box had just been delivered to the door when Lin Yu, without giving them a second glance, was leading his team to analyze the real-time data transmitted back during the rocket recovery.

The screen displayed a dense array of real-time data transmitted from various sensors during the recovery phase. Lin Yu quickly swiped his finger, scrolling down section by section, occasionally pausing at certain points to jot down a few numbers in his notebook.

"At an altitude of 200 meters, sensors 2, 9, 17, and 24 simultaneously went offline, causing the flight control system to lose reliable attitude data." Lin Yu pointed to a timestamp on the screen with the end of his pen. "The last few sets of data received by the onboard computer were distorted values, so the baseline for attitude determination was off."

Soon, the data from the black box was also analyzed.

By comparing the two reports, the team members had already found the root cause without Lin Yu needing to personally explain.

Resonance!

During the rocket's descent, the throttling frequency of several engines and the rocket's structure caused low-frequency resonance. The resonance frequency happened to fall within the natural frequency range of the sensor mounting base, causing the sensors at those locations to first output distorted data and then directly fail and go offline.

The flight control system receives the distortion value, so its attitude corrections are naturally in the wrong direction.

Once the problem is identified, it becomes easier to solve.

Add vibration damping pads, adjust the rigidity of the sensor mounting base, and modify the filtering algorithm of the flight control software.

With these three strategies, you're unlikely to make the same mistake again in the next test.

At the same time, Lin Yu also successfully obtained the two sets of data he was most concerned about.

The first is the temperature distribution curve of the rocket body surface during reentry, and the second is the flight aerodynamic data during the return phase.

He placed the temperature curve side by side with the simulation results from the wind tunnel on the screen, compared them several times, and nodded.

"Impressive, it basically matches the simulation results."

Your special alloy is really tough. Even though the arrow body was blown up like this, the strength of the main load-bearing structures is still well maintained.

"Once the resonance problem is resolved, there shouldn't be any issues with the next recovery."

June 12, Gaolu Town.

Li Ran was preparing to begin his daily routine inspection of the factory area.

The moment he sat down on the electric scooter, Shen Zijie intercepted him.

"Mr. Li, don't turn around yet, come with me to the staff quarters."

Li Ran was driven all the way to the entrance of the residential area. Only after getting out of the car did she realize that Song Xuewu was leading his team to set up something nearby.

Several engineers were squatting next to the green belt, debugging equipment. The communication base station cabinets on both sides of the community gate were all open, and the circuit boards inside had been replaced.

A staff member came over and handed a new phone to Li Ran: "Mr. Li, this is your test phone."

"OK, thanks."

Li Ran took the phone, glanced at it, turned it on and swiped around a bit. Finding no difference from the test unit he had used before, he put the phone away, found a chair, sat down, and quietly watched everyone busy at work.

About fifteen minutes later, Shen Zijie, who had just left, walked out of an apartment building with undisguised excitement on his face.

"Mr. Li, I'm going to show you something amazing today. Hehe."

These words piqued Li Ran's interest.

He dragged another chair over from the side, gesturing for Shen Zijie to sit down, and waved his hand, saying, "Please begin your performance."

Shen Zijie didn't sit down; he stood there, speaking much faster than usual: "Didn't we do a major reconciliation of our technical assets internally last month?"

We found a very interesting document from Su Qing, which is an algorithm framework called "Swarm Distributed Cooperative Control Method".

As you know, the source was the set of technical data that Factory 658 gave us back then.

It had been sitting in the archives for a while because the chip's performance was too poor to run properly.

"It can run now." Song Xuewu, who had just finished debugging a piece of equipment and had a few beads of sweat on his forehead, took over the conversation from the side. "Although the performance of our self-developed 45-nanometer chip is not top-notch, it is cheap, can be produced in large quantities, is robust, and has low power consumption."

This bee colony algorithm is inherently designed for large-scale distributed computing scenarios, making it a perfect match for our mass-produced chips.

"So," Shen Zijie interrupted, gesturing with his hands, "the four of us—Song Xuewu, Su Qing, Zhou Rui, and I—spent some time adapting this algorithm and then porting it to mobile phones, car infotainment systems, televisions, computers, switches, routers, and communication base stations."

We just upgraded the firmware of all the communication base stations and building switches in the community. Now the entire community has formed a shared computing network.

"Give it a try." Song Xuewu pointed to the test machine that Li Ran had just received.

Under Song Xuewu's guidance, Li Ran opened the Lingyun APP.

A new Lingyun SU7 has been successfully bound to the device, and a new icon has appeared in the lower right corner of the interface, which reads "Remote Summon".

He nodded.

A pop-up notification window appears on the screen: "You are currently in a shared computing area. Do you want to enable sharing and call a vehicle?"

Li Ran clicked "confirm".

"Vehicle is autonomously navigating. Please remain within the shared computing area and do not stray far."

About three minutes later, two lights came on from the exit of the underground parking garage.

A Lingyun SU7 with its hazard lights flashing slowly drove out of the ramp and drove at a steady speed along the internal road of the residential area toward Li Ran's location.

There was no driver throughout the entire journey; the steering wheel turned by itself. It automatically slowed down when encountering speed bumps and swerved half a lane to the outside when going around a temporarily parked delivery tricycle. The movements were clean and efficient.

The car wasn't going fast, about the same speed as an adult walking briskly.

Li Ran became interested and turned to walk along the community path as the car approached.

The Lingyun SU7, which had already slowed down and was preparing to stop, seemed to detect that the target location was moving. Its hazard lights came on again, and the wheels slowly turned a small angle, following behind at a leisurely pace.

Li Ran glanced back and laughed out loud: "Not bad, not bad. Others walk their dogs, I walk my car."

He walked around and stood in front of Shen Zijie. The Lingyun SU7 behind him detected the target position and stopped moving, automatically completed the parallel parking, and waited for the driver to enter with its hazard lights on.

The entire sequence of movements was performed flawlessly without any mishaps.

"How did you do that?" Li Ran turned the test machine over and looked at it, then looked up at the base station cabinet on the side of the road.

Shen Zijie squatted down and casually drew a simple diagram on the cement ground with his finger: "The principle is not complicated."

When the Lingyun SU7 needs a lot of computing power to complete autonomous pathfinding and obstacle recognition, its own SoC chip can't keep up.

After all, processing data from six cameras and two millimeter-wave radars in real time, plus the computational demands of the path planning algorithm, is simply too much for a single chip.

"At this point, it will send requests to devices within a certain range that have enabled computing power sharing mode and are currently idle." He drew several interconnected small circles on the ground. "Nearby mobile phones, TVs, computers, and base stations, as long as they have the sharing function enabled and their chips are idle, will contribute their idle computing power to help run the computing tasks required by this vehicle."

Song Xuewu added from the side: "So essentially, within the scope of the community, all idle devices that have enabled computing power sharing are automatically combined into a temporary small computing power pool."

Whoever needs it can use it, and it will be released once the task is completed.

When the vehicle's infotainment system is insufficient, we borrow chips from others, and we also contribute our idle computing power to others at other times.

After listening, Li Ran pondered for a moment: "Hide computing power among the people."

This is an interesting idea.

How is the billing calculated?

"We referenced the time-based billing model used in supercomputing," Shen Zijie stood up, dusted off his hands, "and when your device uses computing cores contributed by others, you are charged 0.01 yuan per core per hour."

"Wait a minute! 0.01 yuan per core per hour? Are you sure that can cover the electricity bill?"

"The actual price is 0.05 yuan per core per hour. The price of 0.01 yuan is the price after our subsidy. After all, there has to be some cost to promote this feature."

Shen Zijie shrugged and said, "If your device is used while it's idle, you can still get paid. The price is the same: 0.05 yuan per core per hour."

We've already conducted internal simulation tests, and for ordinary users, the current cost is actually very low because there are very few scenarios that actually require high computing power. Normally, it would cost at most around 10 yuan per day.

"What about the provider side?" Li Ran asked.

"At this stage, the provider can make a net profit of about two to three yuan per day."

Shen Zijie reached out and wrote a formula on the ground: "If a mobile phone has at least 20 cores for computing, it can earn 20 x 0.05 x 24 yuan a day just by providing computing power, which is at most 24 yuan of passive income."

The initial investment can basically offset the losses. In other words, it's a break-even situation—you get a free, readily available supercomputing quota.

Li Ran was amused by his statement: "Why don't you just say that the income and expenditure are dynamically balanced?"

"That's different." Shen Zijie's expression turned serious. "It's on par now because there are fewer participating devices and the frequency of calls is low."

If this system is rolled out in the future, the number of scenarios in which it is called will only increase, and the daily consumption is expected to rise to twenty or thirty, or even higher.

It's hard to say how much the server will earn at that time; it seems likely that many people will shift from their original earning behavior to simply consuming.

He paused and then said, "And there will definitely be people who specialize in this business and become computing power dealers."

For ordinary users, three or five yuan might not matter, but for those who specialize in buying equipment to build computing pools and then rent them out, it's a different story.

Li Ran thought about it for a moment and felt that what he said made sense.

Currently, the most densely populated public equipment in the entire computing power sharing network is undoubtedly the communication base station. One base station covers an area, and the equipment inside is naturally in the same computing power sharing network.

As data volume and computing demands increase further in the future, each community may need property management to lead the deployment of dedicated high-speed switches and local computing nodes to handle this computing power requirement.

By then, there will be another role for computing power providers.

During their casual conversation, Li Ran had several more conversations with the Lingyun SU7's infotainment system.

Both the accuracy of recognizing continuous voice commands and the smoothness of contextual dialogue have been greatly improved compared to the previous version.

If the old version felt like sending fixed commands via voice, the new version feels more like a smart assistant that reacts quickly and understands perfectly.

"Can this voice assistant be ported to the phone?" Li Ran asked, putting away his phone.

"It's already underway," Song Xuewu continued. "However, in order to be compatible with mainstream operating systems on the market, the mobile applications are actually encapsulated in independent Docker containers."

This model has advantages in security and cross-platform deployment, but the module responsible for permission allocation is not yet properly configured, and we are looking for a solution.