Chapter 413: We’re Just Missing This One Card!
Other experts certainly couldn’t understand the difficulty from a professional perspective, but just relying on intuition, they could tell how hard it was.
Using a typical NP-hard problem with a nearly infinite solution space, those adjectives and the final noun alone make it sound difficult.
Nearly infinite solutions, to industrial experts whose mathematical knowledge is about equal to that of freshman and sophomore mathematics department students, is no different from nearly unsolvable.
The expert sitting next to Wei Zhe could see clues from his expression; seeing Wei Zhe’s stunned look, he knew Lin Ran had produced another earth-shattering result.
Liang Mengsong raised his hand, and Lin Ran nodded: “Engineer Liang, you speak.”
“General Manager Lin,” Liang Mengsong’s tone was heavy, “You have provided us with a solution from the algorithm level, and now we can achieve it in theory.
But to achieve it in the real world, we need a template with sufficiently high precision. With such a template, we can use the NIL lithography machine to imprint the so-called metasurface lens, to have the lens required by traditional route lithography machines.
Let me explain: what we need to manufacture is a metasurface lens master template with 5 nanometer level structures.
And to manufacture this master, we need an electron beam writing machine.”
“This thing is like a pen.” Liang Mengsong tried to explain in the simplest language, “A lithography machine is like a huge printing machine that can replicate the pattern on the template dozens of times in a second.
And the electron beam writing machine is that most primitive pen used to create the printing master.
How fine your pen is determines how fine your master is.
And ours, our domestic most advanced electron beam writing machine, the smallest line width it can stably etch is 65 nanometers.
Using a 65 nanometer pen to carve 5 nanometer patterns is impossible.”
Multi-electron beam writing machine is an advanced semiconductor manufacturing equipment used for high-precision pattern transfer, by using multiple electron beams to simultaneously write patterns on the wafer.
What Liang Mengsong mentioned, the more common term is mask. Light passes through the mask to print the pattern on the silicon wafer, and the precision of the mask determines the precision of the integrated circuit on the silicon wafer.
The machine used to make the mask is called multi-electron beam writing machine; it can only produce one mask at a time, suitable for small batch, high complexity pattern manufacturing needs.
In summary, without the most advanced electron beam writing machine, it’s impossible to manufacture masks for producing the most advanced chips.
It is the “source” of the entire semiconductor manufacturing industrial chain, and its strategic importance is no less than the EUV lithography machine itself.
The world’s most advanced multi-electron beam mask writing machine for EUV level, its market is dominated by Austria’s IMS Nanofabrication company, now absolutely controlled by Intel.
Another Japan company NuFlare Technology is also developing it, but there is a gap in market share and technology application breadth; this is a subsidiary of Toshiba.
Domestic Shanghai Micro Electronics also has a subsidiary responsible for research and development, but the precision is between 90nm and 65nm.
No matter which technology route lithography machine, it needs a mask to print from the master.
In the meeting room, the optimistic atmosphere just ignited by the mathematical breakthrough was instantly extinguished.
Everyone is an industry insider and knows very well that what Liang Mengsong said is fact.
Without the top-tier pen, even the most perfect design drawing is just a piece of waste paper.
Lin Ran said: “This is not a problem. Although we don’t have the most advanced pen, we have the next best pen.”
Lin Ran looked at Yang Depen in the audience and said: “Academician Yang, you speak.”
Then he handed the stage to Academician Yang.
Yang Depen stood up proudly: “Everyone, it’s my honor to tell you that just three months ago, led by our Zhejiang University’s Yuhang Institute, jointly with relevant domestic units, after more than ten years of tackling key problems, China’s first high-precision multi-electron beam writing machine with commercialization capability has completed the final test.
Named after China’s greatest ancient calligrapher — Xizhi.
(China’s first 6nm electron beam lithography machine – Xizhi, from HZ city government official website; don’t know why domestic media didn’t report much, no buzz at all)
“Xizhi is different from traditional lithography machines.” Academician Yang continued: “It doesn’t need a mask, but through computer control, uses a beam of extremely high energy, focused to the extreme electron beam, to directly write circuitry on silicon-based materials.
Therefore, it can modify designs at any time, repeatedly debug, especially suitable for verification in the early stages of chip research and development, and manufacturing unique ultra-high precision components like metasurface lens masters.”
Finally, Academician Yang took a deep breath and announced the two numbers that made everyone hold their breath: “Its performance indicators: minimum stable line width of 8 nanometers, electron beam positioning precision of 0.6 nanometers.”
“Impossible! Positioning precision 0.6 nanometers? Isn’t this precision too exaggerated?”
“0.6 nanometers? Higher precision than the Americans’?”
“Doesn’t that mean we can make 0.6 nanometer lenses? Damn, now the light source can’t keep up with the lens.”
Academician Yang’s face showed undisguisable pride; this was made by our Zhejiang University!
“To help everyone understand, I’ll continue using the pen metaphor.
8 nanometer line width is the stable, clear brush stroke that Xizhi’s pen can write on rice paper.
And 0.6 nanometer precision is the pen tip of this pen, which itself is composed of several even finer weasel hair brushes almost at the atomic scale.
You must have a sharp enough pen tip to possibly control and write extremely fine brush strokes.”
Lin Ran looked directly at Liang Mengsong: “Engineer Liang, I think you should have thought of it.”
At this moment, Liang Mengsong’s eyes were gleaming: “With this 8 nanometer pen, we indeed can’t draw a 5 nanometer master.
The limit of 8 nanometers means we can’t stably and massively manufacture perfect 5 nanometer structures.
Yield rate would be a disaster.
This is a one-time drawing of over 10 to the 14th power antenna structures; it’s indeed impossible.
But we can use this pen to design a smarter brush technique.
We can use DUV lithography machines with multiple exposure technology to make circuit dimensions far below their limit.
Similarly, we can borrow the mature self-aligned quadruple exposure approach in wafer manufacturing, reverse it, and apply it to master manufacturing!
We use Xizhi to first etch a 16 nanometer period base pattern, then through multiple precise depositions and etches, self-grow and replicate a 4 nanometer period structure on the master!
This is extremely complex in process, but completely feasible in theory!”
Lin Ran applauded: “Exactly, I’ve already solved the mathematical level problem for you; master manufacturing needs everyone’s effort. It absolutely won’t be easy, but it can absolutely be achieved.
The final step of imprinting will be done with the NIL lithography machine imported from Canon!
Once successful, relying on China’s manufacturing capability, we will activate an unprecedented 7nm process.
The 28nm process technology transferred to us from Japan during this time, which is fully mature, plus this metasurface lens set, is enough to conquer stable 14nm lithography machines.
And combined with the domestically fully mature double exposure technology, 7nm process is no problem.
If not considering yield rate, combined with quadruple exposure, the limit is 3.5nm.
Already just one step away from the world’s most advanced 2nm abroad.”
Lin Ran’s words caused thunderous applause from the experts in the audience; who dared to think in the past, now we can touch 3.5nm process.
This is not the past pursuit of beautifying semiconductor production lines, but a fully domestic 3.5nm semiconductor production line.
At the same time, it means that once the extreme ultraviolet light source required for EUV lithography machines is conquered, there will be no gap between China and the world’s most advanced 2nm in the lithography machine field.
What does this mean?
The complete failure of the chip war; the chip curtain by America in the past nearly ten years, using every means, uniting all allies, aimed at strangling China’s high technology industry, is thoroughly torn apart and suffers irreparable failure.
Navarro thought he gave China a trap, but it was actually the last card; China holds 2346, just waiting for you to send a 5, Americans personally treat this 5 as a waste card and throw it to the center of the table.
Americans personally treat this club 5 as a waste card and throw it to the center of the table.
This also means that in the future, no allies will cooperate with America to technologically blockade China; this is handing their own industry to China.
The technological north-south war ends; for a long time, the world has been divided into the technology North (led by America) that masters core technology, and the “technology South” responsible for application/manufacturing.
China having 3nm lithography machine means it has completely graduated from the “South” and become a new, powerful, independent kingdom in the “North” camp not controlled by the old king.
From an industrial perspective, once having equivalent manufacturing capability, China will immediately become the world’s third largest, and largest market scale, high-end chip foundry center.
Not only will China’s domestic technology industry explode, once obtaining endless, cheap, most advanced domestic chip supply, China will see explosive growth in all fields dependent on cutting-edge computing power such as artificial intelligence, supercomputing, biotech, autonomous driving.
Cost reduction and computing power liberation will spawn countless new applications and business models.
In such a world, all countries will have to rethink their positioning and future.
The expert scholars present may not think of so much significance, but they absolutely understand that after going through so much hardship, how difficult it is for China, the harder it is, the more important it means, and the greater the final harvest.
The psychological frustration from not being able to go home for the New Year is swept away; now everyone is full of ambition, just waiting to get to work.
Canon Optical Technology Research Institute, Utsunomiya City, Tochigi Prefecture, Japan
“Mr. Ogawa, I understand your difficulties, but commercial contracts are covenants; we must respect the spirit of the contract.”
In the meeting room, the head of China’s import team, Director Liu, had a smile on his face, but his tone was unquestionable.
In front of him, Canon’s vice president, global marketing head Ogawa Kazuhito, had a troubled expression.
“Director Liu,” Ogawa Kazuhito wiped the sweat from his forehead: “We are really doing our utmost already.
Our FPA-1200NZ2C production line is already running three shifts, 24 hours non-stop.
But you added another urgent order of twenty units last month, and require all delivery before the fourth quarter of next year; physically, this is impossible.”
“Nothing is impossible, Mr. Ogawa.” Director Liu calmly took out another document from his briefcase: “This is our latest supplementary agreement; we are willing to increase by 30% on the original contract price, as ‘production capacity rush fee’.
And all payments, within 48 hours after agreement signing, in US dollars, paid in full at once.”
China has scheduled its announcement for the fourth quarter of 2026; currently has 5 FPA-1200NZ2C NIL lithography machines from Canon on hand.
This is also all of Canon’s inventory, including one returned from Hynix, all transported to Shanghai.
All NIL lithography machine production and manufacturing related technology has been taught to China; everything that can be taught has been taught.
But China still adheres to the principle of importing first and then slowly figuring out production itself.
China’s goal is, before it can produce equivalent NIL lithography machines itself, the production capacity obtained from Canon must be enough to last until then.
For Canon, being forced by Washington and Tokyo to sell its soul and future, Canon feels humiliated because this deal is not just selling products, but selling technology.
But the US dollars from China are still very appealing.
At least Canon’s financial report this year will be very good-looking.
Allowing the NIL lithography machine project, which has been unable to profit and merge into financial reports, to bring substantial profit; the entire capital market will look at Canon differently, and if it can capture even part of ASML’s market, Canon’s market capitalization can at least double.
Hynix and Kioxia, these two manufacturers using NIL lithography machines to make storage chips, have pessimistic feedback; Hynix even returned one unit to Canon.
Only China not only doesn’t return goods, but keeps urging them to ship upon arrival, and wants as many as they can get.
Ogawa Kazuhito’s eyes widened instantly.
30% increase, and paid in cash all at once; this is an irresistible temptation for any listed company.
At the same time, Ogawa Kazuhito sighed inwardly: Chinese people still know the goods; Koreans don’t know the goods.
“Additionally,” Director Liu continued to sweeten the deal, “Regarding the supporting nanoimprint resin and template substrate, we know your company’s American suppliers have cumbersome procedures; no problem. Our demand for the next half year is this amount,” he held up five fingers, “five hundred tons; we are also willing to pay 20% premium.
Your company just needs to place the order; all logistics and storage, our own transportation group will send special planes to handle.
You just need to prepare the goods.”
Ogawa Kazuhito was completely speechless.
This is no longer commercial procurement; it’s a near-plundering, cost-unconcerned sweep of goods.
China’s “knowing the goods” certainly excites him, but he can’t understand why Chinese people would place such a crazy bet on NIL, a technology not yet verified by large-scale mass production.
His instinct as a businessman tells him that facing such a “big spender,” the only thing he needs to do is nod, then have the finance department prepare to receive an astronomical sum of cash.
But he still feels uneasy inside, always sensing something is off.
We have researched with Hynix and Kioxia for so many years, and can’t even solve yield rate problems on basic storage chips; can China solve it?
They probably haven’t done production tests yet; where does the confidence come from?
Ogawa Kazuhito has countless doubts in his heart.
