masters-orleans Chapter 193 and other data
Chapter 193 and other data
January 28, 2021.
Shenzhen, Tian Shu chip.
Liu Feng placed the MEMS project evaluation report on He Zhiqiang's desk.
The report is not thick—only eighteen pages in total. The cover reads: "Evaluation Report of TianShu Chip MEMS Project—January 2021".
He Zhiqiang turned to the first page.
The first part of the report is a project overview: TianShu Chip established a MEMS project in early 2020 with an initial team of eight people, which was later reduced to a core team of three. The total R&D investment was approximately seven million RMB, including three million for equipment purchases, two million for personnel costs, and two million for materials and other expenses.
The second part is the technical assessment. Liu Feng used a full six pages to analyze the existing technical accumulation of the TianShu chip MEMS team:
"Our MEMS technology roadmap is based on the traditional second-order thermoelastic coupling correction model, with a process size of 150mm and an accuracy of ±0.045°. This level of technology is above average in the domestic MEMS industry, but there is a significant gap compared with the international first-tier level (Bosch 250mm/±0.030°).
If Wei Lan's third-order model is confirmed, our technological approach will face a fundamental challenge—not a matter of process improvement, but a generational gap in theoretical foundations. Within the framework of the third-order model, our existing second-order modified model will become a special case with lower accuracy.
Part Three provides a detailed analysis of the three exit options:
Option 1: Complete termination.
The MEMS team will be disbanded immediately, and the equipment will be disposed of at a discounted price. The estimated recovery is approximately 1.2 million yuan. Sunk costs are approximately 5.8 million yuan.
Advantages: Fastest hemostasis, resources can be immediately transferred to the MCU.
Disadvantage: If the MEMS industry landscape changes in the future, TianShu Chip will completely lose its opportunity to enter the market.
Option 2: Resell.
The entire MEMS project is being sold as a package—including the team, equipment, and existing technological expertise. Potential buyers include other MEMS startups in China. The estimated valuation is approximately three to five million yuan.
Advantages: Recovers part of the investment while retaining the team (taken over by the buyer).
Disadvantages: Buyers are hard to find. Under the shadow of the third-order model, the market value of traditional MEMS technology is shrinking.
Option 3: Maintain but do not increase investment.
The core three-person team will be retained to maintain basic R&D activities without additional equipment or funding. Annual maintenance costs are approximately 800,000 yuan.
Advantages: Retains the possibility of entry into the market. If future cooperation opportunities arise (such as cooperation with Vilan or obtaining application licenses for the third-order model), TianShu Chip can respond quickly.
Disadvantages: It continuously consumes funds and has no clear output. Investors may be dissatisfied.
After reviewing the three options, He Zhiqiang put the report down.
"Your suggestion?" he asked Liu Feng.
Liu Feng stood in front of the table and remained silent for a few seconds.
"If you ask my technical assessment—" Liu Feng's voice was somewhat low, "Option Three. Maintain."
"reason?"
"If the third-order model holds true, the rules of the game for the entire MEMS industry will change. Everyone will need to relearn and adapt. Whoever still has a team that understands the fundamental processes of MEMS will have the opportunity to find their place under the new rules. If you completely withdraw now, the barrier to entry into the MEMS field in the future will be ten times higher than it is today."
He Zhiqiang leaned back in his chair, tapping his fingers on the table.
"Chen Zheming will not accept Option Three," he said.
"I know," Liu Feng nodded. "An annualized maintenance cost of 800,000 is an 'ineffective investment' for investors. Chen Zheming wants a clear conclusion—to do it or not to do it."
"so?"
"So the final decision depends on how you convince Chen Zheming." Liu Feng looked at He Zhiqiang. "If you can get him to accept the concept of 'observational input'—that 800,000 yuan a year isn't a waste, but a ticket to the game—then Option Three is feasible. If he doesn't accept it…"
"If he doesn't accept, then we'll have to go with option one or option two," He Zhiqiang finished for him.
Liu Feng didn't say anything more.
He Zhiqiang picked up his phone and scrolled through Chen Zheming's chat history. The most recent message was from three days ago: "Has the assessment report been released yet?"
He glanced at the report cover and then pressed the call button on his phone.
"Mr. Chen, the report is out. Would it be convenient to meet and discuss it?"
"Tomorrow afternoon. I'll be in Shenzhen."
"good."
After hanging up the phone, He Zhiqiang turned to the last page of the report.
Liu Feng wrote a "supplementary note" on the last page:
Regarding the review progress of Vilan's paper: According to industry sources, Vilan's NM paper is still under review, with results expected in March or April. It is recommended not to make a final decision before the review results are available. If the paper is accepted, the MEMS industry landscape will fundamentally change, and the value of Option 3 will significantly increase. If the paper is rejected, the decision-making basis for Option 1 or Option 2 will be more robust.
After reading this passage, He Zhiqiang sighed inwardly.
Papers, etc.
Everyone is waiting for the paper.
……
January 29th.
Suzhou, China Resources Microelectronics.
Zhang Li stood in front of the control panel of the 300mm production line, confirming the details of the experimental plan with Chen, the production line manager.
"I've already written the temperature control plan for the 10mm edge area," Zhang Li said, pointing to the parameter table on the screen. "You'll need to make a temporary temperature control modification at the etching station—installing three independent temperature sensors in the edge area to monitor the real-time temperature gradient during the experiment."
"How long will the modification take?" Engineer Chen asked.
"Two days. I've already confirmed with the equipment supplier; the sensors will arrive this afternoon."
"Okay. Will the production line be shut down during the modification process?"
"It will require a four-hour shutdown. Normal production will resume immediately after the modification is completed. The experiment will be conducted during breaks in normal production, without affecting the daily production schedule."
Engineer Chen nodded: "Okay. What kind of cooperation do you need?"
"There are two main things. First, the wafers used in the experiment need to be individually numbered and strictly distinguished from those used in normal production. Second, each set of experimental data requires double confirmation and signatures—one from me and one from your quality inspector."
"No problem." Engineer Chen patted Zhang Li on the shoulder. "To be honest, I'm quite curious about this experiment at Weilan. What exactly does 400mm precision data look like—not just Su Chen, I want to know too."
Zhang Li smiled and said, "You'll know when the data comes out."
"Okay, let's do it then."
Zhang Li turned and walked out of the control room, heading towards the exit along the corridor.
While walking, he took out his phone and sent a message to Su Chen: "Preparations have begun. The sensor arrived today, modification will start tomorrow, and be completed the day after. The first set of experiments is expected to begin on February 1st."
Su Chen replied instantly: "Okay. Data quality is the top priority. Don't rush the process."
"rest assured."
Zhang Li put his phone back in his pocket and walked out of the China Resources Microelectronics building.
The sun was shining brightly outside—a rare sunny day in Suzhou at the end of January.
He stood at the door, recalling the first time Su Chen contacted him a few months ago. Back then, he was just an ordinary engineer at CR Microelectronics, responsible for the daily process optimization of the 300mm production line. Su Chen contacted him because his thesis required SEM data from CR Microelectronics as support.
He is now the fifth author of the paper and is conducting a 400mm verification experiment for a theory that could potentially change the entire MEMS industry.
Life is truly amazing.
……
January 30th.
Su Chen and Zhou Zhiyuan had a long talk.
It's not about peer review—all they can do is wait—but about how to respond to the peer review comments.
"What is the most likely outcome?" Su Chen asked.
Zhou Zhiyuan thought for a moment: "Based on the paper's quality and originality, I think the most likely outcome is a major revision. It's almost impossible for NM to directly accept a 68-page theoretical paper—the reviewers will definitely offer suggestions for revision. The question is, how extensive will the revisions be?"
What changes are most likely to be requested?
"Three directions." Zhou Zhiyuan held up three fingers. "First, some steps in the theoretical derivation need more detailed explanation—our derivation has quite a few steps, and the reviewers may request additional information on the intermediate processes. This is the easiest problem to solve; it's purely writing work."
"second?"
"Second, the completeness of the error analysis. Our error analysis decomposed systematic and random errors, but reviewers may request a more comprehensive sensitivity analysis—such as the impact of material parameter uncertainties on model predictions. This requires additional calculations, roughly one to two weeks of work."
"third?"
"Third, the extrapolation verification of 400mm." Zhou Zhiyuan's expression turned serious. "We gave a theoretical prediction of 400mm in the paper, but without any experimental data to support it. The reviewers will likely request supplementary experimental evidence—even indirect ones."
Su Chen nodded: "So the 400mm edge verification experiment was correct."
"That's right." Zhou Zhiyuan nodded. "Your advance preparation was the right decision. If the reviewers do make this request, we'll already have the data, which will speed up the revision process considerably."
"What about the worst-case scenario?" Su Chen asked.
Zhou Zhiyuan remained silent for a few seconds.
"The worst-case scenario is rejection," he said calmly. "The reviewers believe our claims are too broad and the evidence is insufficient, and they recommend rejection."
What do you think the probability is?
"Not high." Zhou Zhiyuan shook his head. "Even after the controversy surrounding the rejected reviewers, NM's editors still chose to find new reviewers instead of simply desk rejecting the paper—this shows that the editors themselves believe the paper has value. If an editor didn't think the paper had value, they wouldn't waste their time."
"But reviewers and editors don't necessarily share the same judgment."
"Yes. So we can't completely rule out the possibility of rejection." Zhou Zhiyuan looked at Su Chen, "If you are rejected—and I mean 'if'—what are you going to do?"
Su Chen thought about it for a long time.
"I'll submit it to Science or Physical Review Letters instead," he finally said. "Just because NM doesn't want it doesn't mean other top journals won't. The data from the third-order model is solid; any reviewer with a professional eye wouldn't overlook the ±0.018° figure."
"If we resubmit, the review process will start all over again."
"I know," Su Chen said calmly. "If we switch, Lin Wei's 'April meeting point' will be impossible. It might have to be postponed until the second half of the year."
"Can you accept that?"
"It's not a matter of whether I can accept it or not," Su Chen said, looking at Zhou Zhiyuan. "The data speaks for itself. The third-order model is correct. That's a fact. A fact won't change because of a particular journal's peer review results. It's simply a matter of how long it will take to prove that fact."
Zhou Zhiyuan looked at Su Chen, remained silent for a few seconds, and then smiled.
"Do you know who said the exact same thing as you just now?"
"who?"
"Dean Chen," Zhou Zhiyuan said, "when he first saw your third-order derivation last year, his exact words were: 'These formulas don't need me to verify them—the data has already verified them. I only need to confirm that there are no errors in the derivation process.' He also said, 'The data speaks for itself.'"
Su Chen paused for a moment, then smiled.
"Then let's wait for the data."
"Waiting for data."
……
February 1st.
The 400mm edge verification experiment has officially begun.
Zhang Li completed the temperature control modification on CR Microelectronics' 300mm production line, and three independent temperature sensors have been installed. The first experimental wafer, numbered VL-400E-001, entered the etching station at 10:00 AM.
The experimental process is much slower than daily production – every step requires precise control of the temperature gradient, and every set of data requires double confirmation and recording.
Zhang Lishou stood motionless in front of the control panel, staring intently at the temperature curve.
三个传感器的数据在屏幕上实时显示：T1（边缘0mm处）、T2（边缘5mm处）、T3（边缘10mm处）。
Twelve minutes after the etching began, Zhang Li noticed a slight fluctuation in the temperature gradient between T2 and T3—0.15°C higher than the theoretically predicted value.
He frowned, noted down the data point, and continued observing.
The fluctuations disappeared after two minutes, and the temperature curve returned to a stable state.
"记录：T2-T3梯度在第12-14分钟出现0.15°C正偏差，疑似环境温度波动导致。后续需要对比其他实验组数据确认。"张立在实验记录表上写下这段话。
The first group of experiments ended at 3 p.m.
Zhang Li exported the preliminary data and sent it to Su Chen.
After seeing the data, Su Chen made a phone call.
"How does the first set of data look?" Su Chen's voice carried a restrained anticipation.
"It's only the first group, so we can't draw conclusions yet," Zhang Li replied cautiously. "But if you ask my intuition—the temperature gradient distribution in the edge region matches the prediction curve of the third-order model. At least qualitatively."
"What about quantitative analysis?"
"More data is needed. The first group has too little data to be statistically significant. At least twelve groups are needed for reliable quantitative analysis."
"Stick to the plan. Don't rush."
"rest assured."
Su Chen hung up the phone.
He sat in the lab, looking at the first set of data sent by Zhang Li.
The data is still very rough—there is only one set, temperature curves from only three sensors, and only qualitative matching judgments.
But Su Chen saw a signal in these rough data.
A weak, preliminary signal that requires more data to confirm.
But it's heading in the right direction.
……
February 3rd.
Suzhou, Vilan Microsystems.
Production of the first batch of 300mm commercial sensors has entered its third week.
Zhao Guoping inspects the workshop daily, monitoring the production line's operation. The two new Hongyuan machines are performing stably, with the etching station's output about 5% higher than expected—a result of optimized process parameters.
"At the current pace," Zhao Guoping reported at the daily meeting, "the first batch of 150,000 units can be produced by February 18th. Including packaging and testing time, there should be no problem shipping by the end of February."
"Yield rate?" Lin Wei asked.
"93.4%. Consistent with previous batches."
"Okay. Keep it up."
After the meeting, Lin Wei called Zhao Guoping over.
"Any new information from Xingchen Packaging?"
Zhao Guoping shook his head: "No. They haven't contacted us since the last phone call hinting at renegotiating the terms."
"Okay." Lin Wei nodded. "If they call again, tell them the cooperation agreement is valid until the end of March. If they decide not to renew, just follow the procedures."
"Understood. What about Ruicheng Precision and Haina Microelectronics?"
"They didn't say they wanted to renegotiate, did they?"
"No. I just asked about the progress of the thesis."
"Then no special handling is needed. I will provide a unified update on the paper at the quarterly technical sharing meeting—separate sessions for core experts and standard members. Just inform them of the progress; there's no need for excessive reassurance."
"clear."
After Zhao Guoping left, Lin Wei opened her calendar.
February 3rd.
There are still 25 days until the end of February when the shipment is scheduled to be delivered.
There's still some time before the peer review results are out—she doesn't know. Maybe six weeks, maybe eight weeks, maybe longer.
However, the product line is not ready for the paper.
She marked a new date on the calendar: "February 25 - Pre-shipment inspection of the first batch of 300mm sensors".
Then she closed the calendar and opened another document.
This is a document titled "Bosch Follow-up Visit Plan - Draft".
Although she had previously declined Bosch's request for a meeting, she hadn't forgotten about it. Once the paper's peer review results were available—whether acceptance or major revision—she needed to respond to Bosch immediately.
The document states:
Triggering condition: The paper receives peer review comments, and the comments are not rejections.
Response method: Zhou Zhiyuan replied to Stein, indicating that an informal meeting could be arranged.
Attendees (from our side): Lin Wei, Su Chen, Zhou Zhiyuan.
Attendees (Bosch side): Stein + possible Bosch headquarters representatives.
Scope of discussion: Limited to technical exchange, excluding commercial cooperation terms. The goal of the initial meeting is to build mutual trust and understand each other's technological strategies and cooperation intentions.
Bottom line: No disclosure of any unpublished details of the third-order model. No discussion of experimental data for 400mm. No commitment to any form of exclusive collaboration.
Note: We will choose the location for the first meeting—we recommend Weilan, rather than Bosch's Suzhou base. Talking on our turf gives us more control.
Lin Wei read through the draft, made some minor modifications, and then saved it.
She didn't send this document to anyone. Not because of confidentiality—Su Chen and Zhou Zhiyuan both knew about the meeting with Bosch—but because the time wasn't right yet.
Papers, etc.
Everything depends on the paper.
……
February 5th.
Zhihu.
"MEMS Industry Observer" has posted a new thread:
"Weilan's paper has been submitted for six weeks—why haven't we heard anything yet?"
The post was brief: "The paper on the Vilan Microsystems was submitted to Nature Materials on December 25th, and six weeks have passed since then. The review period for NM is usually six to eight weeks. If it were good news, there should have been word by now. Why haven't we heard anything yet? Does this mean there are problems with the paper?"
The comments section exploded again.
"Six weeks is indeed a bit long. But NM's peer review process is slow to begin with."
"Don't worry, six to eight weeks is within the normal range. Wait another two weeks."
"I heard that one of the reviewers at NM refused to review the paper, so a new reviewer was assigned, and the review period started over."
This last comment sparked a wide discussion—some people questioned the source of the information, some analyzed the possible reasons for the rejection, and some asserted that "the reviewer rejected the paper because it was not good enough."
The "calm analyst" did not reply this time.
His silence has made many people uneasy.
"The calm analyst is silent? Does he also think the paper is doomed?"
Don't make wild guesses. They're probably just busy.
"Can discussions on Zhihu determine the fate of a paper? Reviewers don't read Zhihu."
"That's true, but six weeks is indeed a bit long..."
……
Su Chen did not use Zhihu.
He was looking at the fourth set of experimental data sent by Zhang Li.
The four sets of data have accumulated a sufficient sample size for preliminary qualitative analysis. Su Chen opened his mathematical analysis program and overlaid the four sets of temperature gradient data with the 400mm prediction curve of the third-order model.
He stared at the curve on the screen, his eyes narrowing slightly.
The temperature gradient distribution of the four sets of experimental data is in high agreement with the predicted curve of the third-order model.
In the 0-5mm edge region, the deviation between experimental data and predictions is less than 2%.
In the 5-10mm edge area, the deviation is slightly larger, about 4-5% - this may be a system error caused by insufficient temperature control accuracy.
But the overall trend is consistent.
The model predicts that the temperature gradient will exhibit a characteristic "S-shaped" inflection point in the edge region—rising first and then falling—and all four sets of experimental data clearly show this inflection point.
"S-shaped turn".
This feature is absent in traditional second-order models. Only third-order models can predict this behavior.
Su Chen stared at the screen for five minutes.
Then he picked up the phone and dialed Zhou Zhiyuan.
"Professor Zhou, the first four sets of data from the 400mm edge experiment are in. The preliminary results are consistent with the model predictions. In particular, the S-shaped temperature gradient transition in the edge region is visible in all four sets of data."
Zhou Zhiyuan remained silent for a few seconds.
"An S-shaped reversal?"
"Yes. This is a direct reflection of the third-order correction term. The second-order model predicts a monotonically decreasing trend, while the third-order model predicts an S-shaped curve that first rises and then falls. All four sets of experimental data conform to the third-order prediction."
"...This is very valuable." Zhou Zhiyuan's tone changed—from calm to serious. "Su Chen, if subsequent data are consistent, this is not just a bargaining chip to appease the reviewers—it's an independent discovery. The S-shaped temperature gradient transition has never been reported before in the MEMS field. This in itself is worthy of a paper."
"I know," Su Chen said, "but let's focus on finishing the data processing first. We'll deal with the rest after the first paper is approved."
"Agreed. But you need to preserve the data carefully. Each group should have original records, error analysis, and experimental photographs. These will be crucial evidence when writing a second paper or responding to reviewers."
"Zhang Li's team is very meticulous. Each set of data is double-signed for confirmation."
"Okay." Zhou Zhiyuan paused for a moment. "Su Chen, your decision was right. Starting the 400mm experiment ahead of schedule was the right thing to do."
Su Chen remained silent.
"It's not because the data is good that I'm saying you're right," Zhou Zhiyuan added, "it's because you didn't put all your bets on a variable you couldn't control. You created new chips while you waited. That's the right approach."
"Thank you, Teacher Zhou."
"Don't thank me. Keep working."
The phone hangs up.
Su Chen put down his phone and looked at the curve on the screen.
S-shaped temperature gradient transition.
Four sets of data.
Everything matches perfectly.
He said something to himself in his heart—
The signal has appeared.
All that remains is to make the signal stronger, clearer, and more indisputable.
……
February 7th.
Su Chen opened the submission system.
Status: Under Review.
no change.
He closed the page and opened the sixth set of experimental data sent by Zhang Li.
The S-shaped temperature gradient transition at the edge region still exists in the sixth group. The deviation value decreased from 4-5% in the fourth group to 3.2%—Zhang Li made minor adjustments to the temperature control parameters after the fifth group, with significant results.
Su Chen recorded in his notes:
"第6组数据（VL-400E-006）：S型转折确认。T2-T3区域偏差3.2%。温控优化有效。预计第12组之后偏差可以降到2%以下。"
Outside the window is Suzhou in February. The weather is slightly warmer than in January, but it is still gloomy and cold.
Su Chen, wearing a down jacket, sat at his workstation, analyzing data in sets.
Waiting is no longer futile.
New data comes in every day. Each set of data makes the signal stronger.
The paper is in the hands of the reviewers.
The 400mm data is in Zhang Li's hands.
Akira Ishikawa's gyroscope data is in a drawer in Kyoto.
Bosch's evaluation report was on Albrecht's desk.
STMicroelectronics' technical analysis was on Bertoli's computer.
The exit strategy for Tian Shu chips was presented to He Zhiqiang.
The discussion on Zhihu continues to escalate.
Everything—all the lines—is waiting for the same signal.
The results of the peer review for the paper.
And Su Chen—at the center of it all—is doing something simple right now.
Look at the data.
Look at them one group at a time.
Let the data speak for itself.
Let the data prove that the third-order model is correct.
He made data his strongest bargaining chip.
Because when that signal—the peer review results of the paper—finally arrives, he needs to be fully prepared.
Regardless of whether the result is acceptance, revision, or rejection.
He's all ready.
Su Chen saved the data and shut down the computer.
Then he picked up his coat and walked out of the lab.
In the corridor, he ran into Zhao Guoping, who had just returned from the workshop.
"Su Chen, the production of the first batch of sensors is already more than halfway complete. At this rate, finishing ahead of schedule shouldn't be a problem."
"good."
"By the way, Lin Wei asked me to ask you—how's the 400mm experiment going?"
Su Chen thought for a moment and said two words:
"In progress."
Zhao Guoping nodded and did not ask any further questions.
He could tell that Su Chen's eyes were different from a week ago. A week ago, there was anxiety in Su Chen's eyes—the setbacks in the peer review system made him uneasy.
But now, the anxiety has disappeared.
Instead, there is a quiet certainty.
It's like someone who knows the answer but is just waiting for someone else to confirm it.
Zhao Guoping didn't know where this certainty came from—he didn't understand third-order models, 400mm edge verification, or S-shaped temperature gradient transitions.
But he trusted Su Chen.
Because every time Su Chen made a certain decision, it was ultimately proven to be correct.
"Keep it up," Zhao Guoping said, and then left.
Su Chen stood in the corridor, watching Zhao Guoping's departing figure, a slight smile playing on his lips.
Come on.
it is good.
Come on.