Rebirth: Starting from Lighting Up the Tech Tree

Chapter 19 Finals, Show Your Sword

November 17th, Huaxia University Main Auditorium.

The New Talent Cup finals were on a completely different level compared to the semi-finals. The main auditorium, usually used only for school anniversary ceremonies and academician lectures, was transformed into a roadshow venue today—a six-meter-wide LED screen was in the center of the stage, with a row of judges' seats on each side. The first three rows were filled with school leaders, corporate executives, and media reporters. More than three hundred students flooded into the audience area behind, creating a dense, dark mass.

The six finalist teams waited backstage.

Zuo Cheng's team drew the fourth spot to perform. Zhang Lei paced back and forth backstage, silently reciting his presentation script, a pen twirling incessantly in his hand. Chen Hao sat in a corner, giving the demonstration equipment a final check of the connections. Fang Ze leaned against the wall, eyes closed, his breathing even as if he were asleep. Shen Yue squatted on the floor, using tape to secure a loose cable beneath the LED screen—she didn't trust anyone else with it.

Zuo Cheng stood through the gap at the edge of the curtain and glanced out at the audience.

Han Zhe was sitting in the fifth row, by the aisle.

There was another person next to him—Zuo Cheng recognized the square face and frameless glasses.

Zhou Henian also came.

The CTO of Lanwan Communications personally attending a university-level entrepreneurship competition is either out of boredom or there's someone he's after. Zuo Cheng knew which one it was.

He looked away, walked over to Zhang Lei, and patted him on the shoulder.

"Stay calm. Just repeat the same approach as the semi-finals, the only difference is the screen is bigger and there are more people."

Zhang Lei took a deep breath and loosened his grip on the pen. "Brother Cheng, don't worry."

The first three groups' presentations were indeed significantly better than those in the semi-finals. One group's medical AI project had very solid data, while another group's drone swarm demonstration drew gasps of amazement from the audience. Zuo Cheng listened to the applause from backstage, his expression calm.

Strong opponents aren't necessarily a bad thing. It raises the judges' expectations, so if the teams that perform later exceed those expectations, they'll leave a deeper impression.

The fourth group, it's their turn.

Five people walked onto the stage. Shen Yue stood in front of the control panel on the side, Zuo Cheng, Chen Hao, and Fang Ze sat on the technical presentation stage on the right side of the stage, and Zhang Lei stood alone in the center of the stage.

The LED screen lit up.

Shen Yue's opening this time was an upgrade—no longer just a simple video, but a real-time data visualization animation. An aerial view of a virtual city appeared on the screen, showing all the roads, buildings, and base stations. Data flowed between the base stations like blood, with green indicating normal signal and red indicating signal interruption.

As the animation progresses, a virtual car drives into a tunnel, and the surrounding base station signals suddenly turn red. A message pops up in the center of the screen: "Signal lost."

Then the screen split in two—the left half showed the situation without their algorithm, where the signal was continuously interrupted and the red color spread; the right half showed the situation after their algorithm was loaded, where the red color turned back to green within 0.3 seconds and the data flow resumed smoothly.

The contrast was striking. A murmur of discussion arose from the audience.

Zhang Lei's voice followed accurately at that moment.

"Good morning/afternoon, esteemed judges. I am Zhang Lei from the 402 Technology Team. What you just saw was not an animation effect—it was a recreated scene generated based on 72 hours of real-world testing data from a Blue Bay Communication base station using our team's algorithm."

He deliberately paused for a second, letting the keywords "real base station" and "72-hour test" linger in the air for an extra two seconds.

The expressions of the judges in the front row changed. The vast majority of projects in the university-level entrepreneurship competition were still in the laboratory stage; for a student team to obtain actual test data from a real corporate base station was proof of their ability.

Over the next twelve minutes, Zhang Lei streamlined and upgraded his presentation for the semi-finals—focusing on a more specific scenario, providing more concrete data, and clarifying the business logic. Shen Yue's design perfectly complemented his speaking pace and rhythm, while Fang Ze's embedded prototype ran in real-time on the side of the stage, with waveforms on the screen and data on the LED display jumping synchronously.

Zuo Cheng sat on the defense panel and didn't say a word throughout the entire process.

When Zhang Lei finished speaking, the applause was louder than that of the previous three groups.

Q&A session with the judges.

The first person to ask a question was a middle-aged man with extremely short hair, whose name tag read "Fang Mingyuan, Investment Director of Huaxia Guotou". He flipped through the business plan in his hand and went straight to the point.

"Your business plan mentions expanding into radar and the Internet of Things in the medium term. I'd like to ask—what makes you think an algorithm validated in the communications field can be transferred to the radar field? The signal characteristics of these two fields are very different."

This is a probing question. If the answer is not good, the credibility of the medium-term plan will be compromised.

Following the pre-arranged plan, Zhang Lei transferred the call: "Please have our technical lead, Zuo Cheng, answer this question."

Zuo Cheng stood up and walked to the center of the stage.

"Mr. Fang's question is very valid; the signal characteristics of communication and radar are indeed very different. However, our core technological barrier lies not in the specific algorithms—but in the theoretical framework behind them."

He looked at Shen Yue, who understood immediately and switched to a different slide—a diagram illustrating the principle of positive interactive coupling effect.

"During our research and development, we discovered a positive coupling mechanism between modules and rigorously proved its validity. This mechanism is independent of specific signal types; any two signal processing modules that meet the positive complementarity condition can produce a synergistic enhancement effect. It can be used in communications, radar, and even more effectively in low-power IoT scenarios—because the harsher the environment, the greater the enhancement."

He paused for a moment.

"The relevant paper is currently in the peer review process of the Journal of Signal and Information Processing, a top journal in the field of signal processing in China."

Fang Mingyuan paused on the paper, looked up at Zuo Cheng for three seconds, and then wrote a long paragraph in his notebook.

The second question came from an on-campus judge: "Your team of five are all current students. How will you ensure the project's sustainability? What about after graduation?"

Zhang Lei took the question and answered crisply: "Our core team members have reached a consensus—after graduation, we will continue to operate with 402 Technology as the main entity. We are currently applying for incubation at the Huada Technology Startup Incubator, and have also established a formal horizontal research collaboration with Lanwan Communications. The sustainability of the project does not depend on whether we are students, but on whether our technology and customers are real. Our technology has been verified by real base stations, and our customers have paid real money."

This passage was written by Zuo Cheng, but when Zhang Lei said it, it carried a unique confidence and drive characteristic of young people.

Some of the judges nodded.

After the Q&A session, the five people returned backstage. Zhang Lei plopped down in a chair, his shirt soaked through. Fang Ze handed him a bottle of water, then opened one and chugged half of it himself—it was the first time Zuo Cheng had ever seen him drink water so quickly.

Chen Hao adjusted his glasses and, unusually, said something unrelated to technique: "I think we've got this in the bag."

Shen Yue squatted on the ground tidying up the cables, saying without looking up, "Don't milk."

The sixth group—Ma Hao's "Sharp Core Technology"—was the last to perform.

Zuo Cheng didn't return to the audience seats, but instead watched the entire event from a gap in the curtain backstage.

Ma Hao's presentation was an improvement over the semi-finals. He repackaged the presentation logic, turning the AI ​​model training process into a visual animation with visual effects comparable to Shen Yue's design. His technical explanations were also more polished, and he prepared responses to several questions he had previously avoided.

But Zuo Cheng heard a flaw in the clue.

During the Q&A session with the judges, Fang Mingyuan asked Ma Hao a question: "How well does your AI model generalize in scenarios with few samples? How do you guarantee the reliability of the model if there is insufficient training data?"

Just as Zuo Cheng predicted.

Ma Hao was clearly prepared for this question, and answered calmly: "We used transfer learning and data augmentation techniques to address the small sample size problem. Experimental results show that when the training data is reduced to 30%, the model accuracy loss is no more than 8%."

The answer wasn't wrong, but Zuo Cheng shook his head.

What does an 8% accuracy loss mean in the communications field? It means that out of every 100 channel estimations, eight are inaccurate. In boundary scenarios such as high-speed movement or low signal-to-noise ratio, these eight inaccuracies can lead to call interruptions or data packet loss. Industrial-grade systems typically tolerate accuracy loss of less than 3%.

Ma Hao didn't mention this threshold, and the judges didn't press for details. But Zuo Cheng knew that Fang Mingyuan must have already calculated it in his mind.

After all the presentations were completed, the judges left the room for discussion.

Zuo Cheng returned to the audience seats and found someone sitting next to him—Yu Ying.

"When did you arrive?" He was a little surprised.

"They arrived at the beginning of the third group." Yu Ying looked at him, a look in her eyes that Zuo Cheng rarely saw on her face. "I've watched your entire roadshow."

"How is it?"

Yu Ying was silent for a second or two, then said two words: "That's good."

This isn't an evaluation of the technology, nor is it an evaluation of the business plan. It's just two simple words—"Great."

Zuo Cheng paused for a moment, then smiled.

"Thanks."

Twenty minutes later, the host went on stage to announce the results.

There were three groups that won third prize. After the names were read out, 402 Technology was not among them. Zhang Lei clenched his fists.

The second prize winners were from two teams. After their names were read aloud, it was read: "Sharp Core Technology, Ma Hao's team."

Ma Hao stood up and walked onto the stage to accept the award, his expression composed as he smiled and shook hands with the judges. But as he stepped off the stage, his gaze swept over the direction of Zuo Cheng, lingering for only a fraction of a second.

Zhang Lei's breathing quickened, and Liu Wei sent more than a dozen question marks in the group chat.

"First Prize—" the host opened the envelope, "402 Technology, Zuo Cheng Team!"

Zhang Lei jumped up from his chair, Chen Hao's glasses nearly flew off, and Fang Ze's lips curved into a smile Zuo Cheng had never seen before. Shen Yue let out a soft scream from the technical area, then immediately covered her mouth with her hand.

Zuo Cheng stood up and walked onto the stage.

When the lights shone on him, his gaze passed over the judges' table, over the audience, and over three hundred young faces, landing on the last row near the door.

Han Zhe and Zhou Henian were sitting there. Zhou Henian was clapping, his face expressionless, but his clapping was very forceful.

Zuo Cheng accepted the trophy and certificate, and said only one sentence into the microphone.

"Thank you to the judges, and thank you to everyone on the team. This award belongs to all members of 402—including the one who isn't on stage today."

In a corner of the audience, Liu Wei took off his glasses and rubbed his eyes hard.