signingbooks Chapter 246 The "Big and Small Hearts" on the Second Floor of Enlightenment
Chapter 246 The "Big and Small Hearts" on the Second Floor of Enlightenment
Beijing, Zhongguancun.
While almost all the lights in Zhongguancun's tech companies had been turned off, the research department floor of the building at BGI Jiutian's headquarters remained brightly lit.
The security guards downstairs have already changed to their second shift.
Upstairs in the research department, dozens of engineers were still typing away at their desks, the characters scrolling down the screen line by line.
It's only a little past 8 p.m. now.
But for the engineers at BGI Genomics, 8 a.m. was just the beginning of the second half.
It's common for people to voluntarily return to the company to work overtime after dinner.
no way.
The research funding from above is far less than that of Synaptics, Cadence, Ansys, and Siemens—it's less than an order of magnitude. But the tasks undertaken by BGI Nine Sky are exceptionally difficult—
They want to create an industrial simulation EDA system that truly belongs to China.
We need to tear a hole in that web woven by four oligarchs for over twenty years.
The goal is to one day allow Chinese design companies and contract manufacturers to no longer have to bow down to Synaptics.
Insufficient research funding and insufficient research personnel—then make up for it with overtime work.
Over the years, many people couldn't stand the daily overtime, the endless work, and the low wages, and gradually left Huada Jiutian.
But more people stayed.
They are willing to work overtime voluntarily, even if the overtime pay is very limited.
They stayed because they wanted to see the day when Chinese industrial simulation EDA truly went global.
These are the most lovable people at Huada Jiutian.
"Mr. Hu, it seems Su Chen is going to fail this time."
In a small conference room in a corner of the research department, several engineers were sitting around a round table, discussing in hushed tones the Bloomberg Surveillance news article that had been widely circulated in major tech media outlets that day.
The speaker was an engineer in his early thirties. He looked at the 53-year-old man opposite him and expressed his opinion.
This 53-year-old man is named Hu Zhiqi.
One of the founders of Huada Jiutian, and currently the company's strongest core figure in terms of research capabilities. He holds a PhD in Microelectronics from Beijing University of Posts and Telecommunications, graduating in 1995. From that year onward, he dedicated himself to the domestic EDA field for a full thirty years.
But before Hu Zhiqi could answer, another engineer next to him spoke first.
"But this is fine too," he said. "Since the Sentaurus system can't accommodate us, and since Boston won't let global contract manufacturers sign contracts with us—then we'll work with Vilan to create a brand new 'Enlightenment System' and build a new EDA empire."
This statement sounded impressive, but the young engineer wearing glasses opposite him gently shook his head.
"I'm afraid it won't be that easy." The engineer with glasses sighed. "Those thousands of industrial simulation application companies around the world are all money-grubbing. I doubt they'll invest money to develop a new version specifically for Vilan's Qiming Layer 2 interface specification."
The group fell silent immediately.
This is the heaviest stone in the hearts of the engineers at BGI Jiutian.
This rock is very difficult to move.
Even if Vilan can indeed launch the world's most powerful industrial simulation OS, without the support of those thousands of application companies and those thousands of industry simulation tools—automotive, aviation, biomedicine, energy, and semiconductor design itself—then Qiming Level 2 will just be a castle in the air, and no one will be willing to climb up.
Even if the contract manufacturer is willing to sign off, it's useless.
Because there are no applications, the design companies will never use the layout approved by the OEM to make products.
"I don't know—" The engineer, who was in his early thirties and spoke first, turned to Hu Zhiqi. "What does Mr. Hu think?"
Several pairs of eyes turned to look at them.
Hu Zhiqi slowly raised his head.
He didn't answer immediately, but instead picked up the cup of tea on the table that had gone cold, took a small sip, and then gently put the cup down.
"This time, Weilan did not obtain the compatibility certification from the Boston Global IP Council," Hu Zhiqi said slowly. "From a certain perspective, this seems to be a good thing for us at Huada Jiutian."
Several people looked up.
"Having a powerful company like Vilan willing to work with us to support the Qiming Layer 2 interface specification and maintain a 'Chinese system' is indeed a good thing for us."
Hu Zhiqi paused.
"But this is an unfortunate thing for the entire Chinese industrial simulation industry."
He let out a long sigh.
"Because without Boston's compatibility certification, it means that Qiming's second-layer simulation output cannot be directly approved by the four top global contract manufacturers."
"Being unable to run tape-out and approval processes at TSMC, Samsung, GlobalFoundries, and Fischer Semiconductor," Hu Zhiqi said, emphasizing each word. "This means our industrial simulation OS is destined to be a 'niche OS.'"
Silence fell over the meeting room once again.
Hu Zhiqi was actually more upset than anyone else.
Villon spent $3 million on the Sentaurus compatibility license, but after that one sentence from Boston, it turned into a joke that seemed almost like "buying nothing at all".
The door to the path Su Chen was about to take was slammed shut from the outside.
This path is not only Su Chen's path, but also Hu Zhiqi's own path, the path that Huada Jiutian has walked over the past thirty years, and the path that the four Chinese domestic industrial simulation companies—Huada Jiutian, Gailun, Guangliwei, and Xinhuazhang—have wanted to take over the past twenty years.
Everyone on this road saw the door being closed.
Everyone present understood this sense of sorrow.
In the midst of the silence, a middle-aged executive who looked somewhat greasy slowly spoke.
He is the CFO of Huada Jiutian, and his surname is Zhou.
"Mr. Hu," Mr. Zhou frowned, "Weilan's side—someone came again yesterday."
Hu Zhiqi looked up.
"They want to directly buy out the subsequent patent fees for the **'Local OEM Sign-up Protocol Adaptation Layer Based on Qiming Layer Interface'** that we developed this year."
The eyes of several young engineers immediately turned to Hu Zhiqi.
They all knew that this adaptation layer was the result of almost everyone at Huada Jiutian working overtime throughout the year. Its core value was to translate the interface specifications of Qiming Layer 1 into a format that could be directly approved by the five domestic foundries—CR Microelectronics, SMIC, JCET, Huatian, and AVIC MEMS.
It's not just a simple gadget.
It is a localized approval agreement that BGI Nine Days developed from scratch in one year, capable of challenging the Sentaurus approval process.
Vilan wants to buy out the remaining patent fees.
"President Hu," President Zhou paused for a moment, "what should we do?"
All eyes turned to Hu Zhiqi.
Hu Zhiqi did not hesitate for long.
"Sell it to them at a low price," he said.
President Zhou frowned.
"Mr. Hu, this—"
"According to the agreement signed by both parties at the time," Hu Zhiqi explained slowly, "Weilan was originally entitled to automatically obtain the license to use, adapt, and further develop this adaptation layer that we developed based on the Qiming Layer 1 interface. They just don't want to continue paying patent fees annually in the future; they don't want to buy the ownership of our adaptation layer."
Mr. Zhou nodded.
"Moreover," Hu Zhiqi continued, "if Weilan really adopts our adaptation layer and implements it in Qiming's second-layer OS, it would actually be a good thing for us."
"Why?" a young engineer couldn't help but ask.
"Because our adaptation layer will be able to directly enter the market with the help of Vilan." Hu Zhiqi smiled. "If Huada Jiutian were to promote it on its own, it might take ten years to get factories other than local OEMs to use it; with Vilan's promotion, this system can be directly deployed to every machine of Qiming's Layer 2 OS."
"In a unique way, we achieved a domestic EDA substitution."
The engineers looked at each other.
"Furthermore," Hu Zhiqi's tone gradually turned somber, "what does Wei Lan have? They have the Qiming Big Engine, the Qiming Layer 1 interface specification, the Qiming Layer 2 OS, 180 MEMS algorithm patents, the Sentaurus compatibility license from 1999, Magwel, and the old-school algorithm foundation passed down by Mr. Shen Shouhong."
“They have,” Ho Chi Kee looked up, “far more than we do.”
"If we don't satisfy Vilan, they can start from scratch and develop a completely new local OEM signing agreement adaptation layer, and completely abandon us."
"By then, all the results we've achieved through a year of overtime work," Hu Zhiqi's voice was flat, but every word was heavy, "will just be a useless joke."
Another silence fell over the conference room.
This silence carried a kind of clarity that came after being awakened.
"For China's industrial simulation to surpass Synaptics, it must contend with the vast network of Boston Dynamics-Synaptics-Foundries-Application Layer," Hu Zhiqi concluded, adding, "The four of us—Huada Jiutian, GILUN, GLONASS, and Chipmax—must maintain a consistent stance."
"We cannot waste our research resources on developing two incompatible interface standards. That would further scatter our already limited resources."
"Follow Qiming," he said, emphasizing each word, "Follow Qiming's Layer 2 Interface Specifications."
President Zhou frowned, still somewhat hesitant.
"But this is the culmination of a year's hard work. Selling it at a low price, and even in the form of buying out all subsequent patent fees..." Mr. Zhou rubbed his hands together. "If we really do this, we'll never receive patent fees for this line again. Just thinking about that money—"
He didn't finish his sentence.
"President Zhou," Hu Zhiqi looked at him, "don't forget our purpose."
"Don't forget why we pooled our money to start this company in the first place."
"Don't forget the nights we've spent together over the past thirty years."
President Zhou quieted down.
Compared to this greasy CFO, Hu Zhiqi clearly understands better what they want.
Several engineers in the conference room nodded in agreement.
This year's hard work is real.
But as long as China's domestically developed industrial simulation EDA can reach a point where it "runs alongside the world," what does it matter if subsequent patent fees for this line of technology are no longer collected?
And more importantly—
They remain the original developers of this adaptation layer.
This is equivalent to Intel holding the x86 instruction set itself within the x86 architecture.
Vilan only bought out the "subsequent patent fees", not the "original status".
Huada Jiutian, as the first company to translate the Qiming Layer 1 interface into a local outsourcing approval process, will always hold this position.
certainly--
Wei Lan may not necessarily adopt the entire adaptation layer system that Hua Da Jiu Tian wrote in a year.
More likely, Vilan will select a portion of the interface modules that it actually needs, and then have its own engineers rewrite the rest, ultimately assembling a completely new "local signature protocol layer" belonging to Qiming Layer 2 OS.
If this is done, BGI Jiutian's share in that newly signed agreement layer may only be a large portion, or it may only be a small portion.
BGI Jiutian will never become the "Chinese version of Intel".
but--
BGI Nine Sky can at least become the "Chinese version of Kentech".
Enter the inner circle of Qiming's big engine ecosystem.
Obtained production and deep integration licenses for Qiming Layer 2 OS.
It became the first of the four domestic industrial simulation companies in China to officially launch its business.
This is the saying, "When one attains enlightenment, even his chickens and dogs ascend to heaven."
The engineers in the conference room exchanged glances, and their eyes gradually lit up.
Hu Zhiqi picked up the cup of cold tea and drank the last sip.
"Tomorrow morning," he said calmly, "I will go to Shanghai myself."
Shanghai, Pudong Zhangjiang South Area, Sunqiao Building No. 3, 6th Floor, East Unit.
Chen He stood at the door of Su Chen's study.
Only one desk lamp was lit in the study.
Su Chen sat behind his desk, elbows propped on the surface, fingers interlaced under his chin, his gaze fixed on a sheet of paper spread out on the table—at the top of the sheet were five words:
Qiming Two-Tier OS Architecture Diagram (Draft v0.7)
Chen He has just finished his report.
During this period, he reported in detail on a series of actions by Synopsys, the Boston Global IP Committee, Ansys, Siemens EDA, and Kentech.
He reported in great detail on dozens of companies, hundreds of journalists, hundreds of media headlines, and thousands of social media discussions, both domestic and international.
He has already reported Hu Zhiqi's attitude from Huada Jiutian.
After he finished speaking, he stood there, waiting for the boss's reaction.
Su Chen didn't speak.
His fingers tapped lightly on the table.
"Da-da-da-da-da-"
Knock in all directions.
Hold for a second.
Tap four more times.
Chen He did not disturb him.
He knew that Su Chen shouldn't be disturbed when he was thinking. In the past year, he had witnessed too many of these silent knocks—each time, the words Su Chen spoke afterward would lead Wei Lan down a path that no one had ever thought of before for the next three months, six months, or a year.
The lamp cast a long shadow of Su Chen onto the wall of books behind him.
a long time--
Su Chen finally stopped pointing.
He raised his head.
"Chen He".
"exist."
"Since Boston can't accommodate us," Su Chen said, enunciating each word clearly, "then starting today, we'll simply give up on begging Boston for compatibility certification."
Chen He paused for a moment.
"We directly use the Qiming Big Engine as the core and support the Qiming Layer 2 OS on its own. This ensures perfect compatibility between the OS, engine, interface specifications, and local signing protocol layer, maximizing performance," Su Chen said.
Chen He frowned.
He had also been cramming on a lot of knowledge about industrial simulation EDA during this time, and he knew—
"But then," Chen He said cautiously, "wouldn't that mean we wouldn't be able to use those industrial simulation applications in the Sentaurus ecosystem?"
He knew that in order for design companies, car manufacturers, aerospace manufacturers, biopharmaceutical companies, and semiconductor design companies to use Vilan's Qiming Layer 2 OS, they had to be able to run their existing simulation tools—car crash simulation, fluid dynamics, chip signal integrity, and MEMS coupling—on Qiming Layer 2.
Most of these tools are written based on the ecosystems of Sentaurus, ANSYS, and Kenteng.
If we abandon the Sentaurus ecosystem directly, it means that all these tools will suddenly become unusable overnight.
Who still uses Qiming's Layer 2 OS?
His confusion is quite reasonable.
But for Su Chen—
This is just a small matter that I've already thought through.
Su Chen smiled.
"Chen He, don't forget," he said, "that Sentaurus mini-engine that our Qiming two-layer processor is working on."
Chen He looked up.
"We can develop a 'signature-compatible container' in the Qiming Layer 2 OS," Su Chen said, emphasizing each word. "In the form of a container, we can directly run industrial simulation tasks that could originally only run in the Sentaurus ecosystem."
His words were like a lamp, suddenly illuminating the chaos in Chen He's mind.
Chen He's eyes suddenly lit up.
He got it.
This tactic—
To put it bluntly, it's essentially the same trick that Guowei Technology Company played with mobile operating systems starting in 2019.
The problems that Guowei Technology faced back then were:
HarmonyOS is a completely new system; it has no games, no software, and no applications.
Without games, software, and applications, this system won't be accepted by consumers. Even if Google forces HarmonyOS onto its own phones, consumers will simply flash it back to Android. Or worse—consumers will simply abandon Google phones altogether.
As a completely new system, HarmonyOS must have "enough available applications" on day one.
What should I do?
The country has come up with a clever idea regarding science and technology.
Since we don't have our own application, let's use someone else's first. Let's solve the "having or not having" problem first, and then we'll worry about the rest.
Therefore, Guowei created a software-level virtual container within the HarmonyOS system, enabling HarmonyOS to directly run games and software from the Android system.
This move directly reversed the situation where the HarmonyOS system was "destined to be unused".
With this strategy and the huge existing user base of its own mobile phones, the number of HarmonyOS installations officially exceeded 3 million in the second half of 2024.
3 million installed units is the best GG (Golden Goal).
With 3 million units installed, that's the best leverage.
After that, everything went exactly as expected.
Several domestic and international game companies and application companies have proactively come to us to develop native versions specifically for the HarmonyOS system.
HarmonyOS successfully reversed its unfavorable situation by using the strategy of directly running Android applications through "software-level virtual containers," making its system the world's third largest operating system.
Su Chen is now using an industrial simulation version of this move.
and--
Chen He gradually understood why Su Chen had insisted on the "big heart, small heart" structure since last December.
Su Chen is not trying to create a "software-level virtual container".
What Su Chen needs to do is—
Hardware-level "signature-compatible container".
He wants to reserve a dedicated Sentaurus small engine hardware core on the Qiming Layer 2 processor, as well as a matching "signature data path" hardware IP, to provide a good path for this container at the hardware level.
The performance loss of software-level virtual containers can be as high as 30%–50%.
Performance loss of hardware-level sign-off compatible containers —
Vilan's internal testing showed a result of approximately 5%–8%.
This is a difference that cannot be compared to the past.
Of course, the development cost of hardware-level containers is much higher than that of software-level containers.
But Su Chen has clearly decided to bear this cost.
Chen He looked up at Su Chen.
The table lamp divided Su Chen's face into two halves, one light and one shadow.
Chen He noticed that his palms were sweating.
He had heard it too many times: Su Chen would reveal a hidden move that outsiders couldn't understand at crucial moments.
But this time—
This move is to add another layer of protection to the "compatibility certification" system that the Boston Global IP Council has carefully maintained for over two decades.
No, it's not about adding a step.
Instead, they removed the door from the inside.
"Boss," Chen He took a deep breath, "if we really go ahead with this—"
"Boston won't let this go," Su Chen said calmly. "Neither will New Thinking."
He raised his eyes and looked at Chen He.
"But what can they do?"
"Qiming Layer 2 OS is an open-source OS." Su Chen said, word by word, "The code for Sentaurus-signed compatible containers will be submitted to the Qiming Layer 2 OS GitHub repository by 'amateur experts' in the Chinese EDA community in the form of anonymous patches."
"Weilan—" Su Chen smiled, "It's just a PR merger."
Chen He's breath hitched.
He finally understood what Su Chen had been tapping on his desk that night.
Hardware-level signature compatible container.
Open source OS.
Chinese EDA community unofficial patches.
Three moves in quick succession.
Boston wants to sue—
Whom can it sue?
It sued the entire Chinese EDA engineer community?
It sued thousands of microelectronics graduate students from Beijing University of Posts and Telecommunications, Tsinghua University, Fudan University, Shanghai Jiao Tong University, Huazhong University of Science and Technology, University of Electronic Science and Technology of China, University of Science and Technology of China, and Xidian University?
It sued thousands of engineers from more than 20 local EDA startups in Zhongguancun?
It cannot be persuaded.
It can't be sued by anyone.
Qiming's second-layer OS will gradually move the global industrial simulation application ecosystem under Qiming's big engine in a way that is like "opening a new window in the wall when the door is closed".
Move for a year.
We moved two years ago.
Move to the first quarter of 2028, the day the Qiming Layer 2 Interface Specification is officially released and launched.
By then—
Looking back, the Boston Global IP Council realized that the tightly locked door to "compatibility certification" had become a door that nobody cared about.
Chen He looked at Su Chen under the lamplight, at the drawing labeled "Qiming Two-Layer OS Architecture Diagram (Draft v0.7)," and a sentence suddenly popped into his mind—
Older gingers are more spicy.
But he immediately shook his head.
wrong.
Su Chen is only 28 years old.
He recalled the 45-year-old algorithm engineer in the tavern who, in silence, uttered the words, "I think Su Chen can do it."
That night in Shanghai—
Hundreds of engineers from Qiming Layer 2 OS were laughing heartily in Building 3 of Sunqiao after figuring out the key to "open source OS + community patches";
Several engineers from BGI Genomics, in a small conference room in Zhongguancun, Beijing, gradually had a glimmer of hope in their eyes as they figured out their position as the "Chinese version of Kentech."
And Su Chen—
Under the desk lamp, he gently rolled up the architecture diagram with his fingers and tucked it into the third drawer on the right side of the desk.
Inside the drawer lies a letter from Mr. Shen Shouhong dated 1995.
The card indicating that Weber met in Shanghai on August 12th was pressed down.
The list of five directions for Qifeng Academy, dated July 7th, is pressed down on the page.
Now, we need to press on—
This is a "Qiming Two-Layer OS Architecture Diagram (Draft v0.7)".
And that night, Su Chen personally drew two small characters in the lower right corner of the drawing and signed them with a blue pen:
"Big Heart" and "Small Heart".