Tom had originally thought that, with the successful breakthroughs in 7 nm, 5 nm, and 3 nm technologies, the 2 nm chip technology would also be successfully overco.
However, he did not expect that at this mont, closest to the limit, nurous extrely complex and difficult technical problems would all erge.
"Thinking about it, this is actually quite normal."
Tom sighed softly: "To precisely engrave over 40 billion transistors onto a silicon wafer within a few square centiters, and to make them maintain the connections from the blueprints, working together to function—how could that be simple...?"
But no matter how difficult, it had to be done.
Tom sorted through the existing technical obstacles and once again began the breakthrough mission.
He discovered that a major constraint preventing him from mastering 2 nm chip technology was a more severe short-channel effect compared to previous generations.
Leakage current was almost uncontrollable at such minuscule dinsions. This led to transistors being unable to function properly; once powered on, a large number of transistors would instantly burn out.
Tom mobilized the brainpower of tens of thousands of clones, simultaneously thinking, using supercomputers for simulations, and conducting various experints.
Finally, he conceived a solution.
"Perhaps... I can change the structure of the transistor? By wrapping the channel with the gate, I might be able to improve control and reduce leakage current."
Upon this thought, Tom imdiately began to experint.
Using the high-precision lithography equipnt in the laboratory, Tom attempted to create a brand-new chip based on the gate-all-around field effect mode and powered it on for testing.
After a series of tests, looking at the current data on the screen, Tom let out a long breath.
He knew that the technical problem of leakage current had been overco.
However, this was only the first difficulty encountered in the process of developing the 2 nm chip.
After this, Tom imdiately encountered the second technical challenge.
How to further increase light source power to achieve a shorter wavelength?
Clearly, only with a shorter wavelength of light could smaller transistors be engraved.
But currently, the lithography machine Tom was using could already be called an "extre ultraviolet light source" lithography machine.
The "extre" in it ant extre.
But now, even "extre" was not enough; a way had to be found to further improve it.
Tom once again gathered a large amount of brainpower and resources, beginning cost-unrestricted attempts.
After several days, a new idea erged from the mind of one of the clones and was synchronized into Tom’s brain.
"Perhaps other materials could be tried in the laser plasma light source."
After a series of attempts, Tom finally discovered that bombarding tin droplets with a high-power carbon dioxide laser generator could produce a light source with a wavelength that t his requirents.
This problem was also overco.
However, with the light source problem solved, the photoresist problem surfaced.
Existing photoresists could not effectively absorb and react with this wavelength of light, making it impossible to effectively engrave transistors on silicon wafers.
Furthermore, its sensitivity was not high enough, and while its related defects were acceptable in the previous generation of chips, they were magnified to an unacceptable degree in the manufacturing of the 2 nm chip at this mont.
"New materials need to be developed."
Tom pondered, resuming material science experints, directing hundreds of thousands of clones to conduct tens of millions of experints in total. Finally, he found a material that t the current requirents.
This was a new type of photoresist based on tal oxides and organic-inorganic hybrid materials, which perfectly solved all the problems of the previous generation of photoresist.
After this, new problems continued to erge.
The insufficient reflectivity of the mask used in the lithography process led to energy loss from the light source, making it impossible to engrave transistors of sufficient quality.
Tom once again launched a concentrated effort, attempting to use multi-layer molybdenum and silicon reflectors, which successfully resolved this issue.
After this, one problem after another still appeared.
Problems related to the optical system, stability, precision issues, thermal managent, cooling systems, and so on—the problems were almost endless.
Often, one problem was solved, and it was thought that progress could be made smoothly, but the next mont, hundreds or thousands of problems would appear.
After solving these hundreds or thousands of problems, tens of thousands more would then appear.
Sotis, an extrely minor issue could tie up millions of the chip research clones under Tom for over ten days. But even if Tom mobilized imnse brainpower and material resources to finally solve this problem, he would find that it was still only a small step forward.
Ti slowly passed amidst these constantly erging problems and the ceaseless attempts, thoughts, and research of millions of clones, day and night.
Finally, one day, Tom solved another problem, and when he subconsciously looked at the problem list, he was startled to find that the problem list... was empty.
Empty?!
A wave of excitent instantly surged through Tom’s heart.
After such a long period of struggle, had he finally mastered 2 nm chip manufacturing technology?
Although what he had mastered at this mont was only the manufacturing technology in the laboratory, there was still so distance to go before mass production.
Once the mass production stage was entered, more problems would inevitably erge, requiring him to solve them one by one.
But at least, from the perspective of basic principles and processes, he had truly mastered this technology! All subsequent problems would only be technical problems, no longer problems of principle!
In his excitent, Tom imdiately began to upgrade and transform one of the factories that originally produced 3 nm chips, manufacturing more precise equipnt, replacing existing equipnt, and attempting trial production.
As soon as trial production began, sure enough, nurous problems resurfaced one by one. But it didn’t matter; Tom had a million clones on standby, ready at any mont.
Problems appeared; they just needed to be solved.
Thus, this chip factory began rapid optimization and iteration. Finally, after half a year, the production line stabilized.
The first batch of 2 nm process chips was finally produced.
Although the yield rate at this mont was less than 10%, aning that out of the factory’s annual production capacity of 10 million chips, less than 1 million chips t production expectations, it didn’t matter.
At least the supercomputer could be built first! The yield rate could be slowly improved later.
Therefore, all of the factory’s good products for the first full year, approximately 1 million chips, were taken by Tom and used entirely in the construction of just one supercomputer.
At the sa ti, a full 100 newly constructed 2 nm process chip factories had already begun production.
Reviews
All reviews (0)