"It's the data that's wrong, not our calculations!!" Boss Jia licked his lips as an idea became clearer in his mind. Finally, he asked directly, "Teacher, where did the basic data we used to calculate come from?"
"Of course we measured it ourselves!" Ni Guangnan replied subconsciously.
"That's not right. Some of the data weren't measured at all. I remember that there was a part of the data that I was responsible for checking and calculating. It didn't seem to be the result of measurement!" Boss Jia squinted and said.
"Okay, this is the data that we directly plucked from the ASML machine!" Ni Guangnan said somewhat embarrassedly, "In order to save time, I directly plucked the data from the ASML 70nm machine!"
"But I guarantee that this part of the data is universal for both dry and immersive methods. There's no impact at all. And this data can't be measured at all. It needs to be accumulated through countless experiments! "
"For example, the problem of the engraved polarization angle. It can't be measured at all. It can only be measured from 44.0000001 degrees to 45.9999999 degrees. Every 0.00000001 degrees can be experimentally calculated to find the best polarization angle."
"If we rely solely on experiments to find it, we'll have to perform 20 million calculation experiments. Just calculating a polarization angle will take more than three years. Where do we get that much time?"
"And the polarization angle is the same for both dry and immersive methods. There's no impact at all. That's why I directly used ASML's polarization angle data! Anyway, it's all 70nm and the data is the same. We don't need to spend a lot of time experimenting because the final result of the experiment will be the same as the ASML. It's too wasteful! "
Ni Guangnan explained his behavior. Although copying the data sounded a little shameless, it was actually very necessary.
Because a lot of data couldn't be calculated or measured, it had to go through tens of thousands or even tens of millions of experiments. It was better to directly copy and paste from an existing object.
For example, one of the classic arguments in Hua Xia's essay, Edison's Creation of the Light Bulb.
It was impossible to say whether this story was true or not. Assuming that this story was completely true, Edison had experimented for a full 13 months in order to find the most suitable material to make silk. He had tried more than 6000 materials before he finally found bamboo silk.
What's more, it wasn't ordinary bamboo. It had to be burned in a high-temperature sealed furnace and then processed to obtain charred bamboo filaments. Only then could it be placed in a light bulb.
Edison had spent so much effort to find the material. Back then, he only needed to buy an incandescent light bulb, take out the filament inside, and reverse engineer it. In less than two days, he knew what the best material was.
Ni Guangnan was doing the same thing. Because the collection of experimental data was too time-consuming and expensive, he might as well buy an ASML 70nm lithography machine and obtain the data directly. It was a lot faster.
Although this method was very shameless, it was very effective. Otherwise, Boss Jia would have to wait at least three years before he could see this experiment in front of him.
This method was also approved by Boss Jia. Otherwise, why would Boss Jia spend so much effort to buy two of the most advanced 70nm lithography machines? It was all for Ni Guangnan to reverse engineer it and obtain the data inside.
This was something that Boss Jia and Ni Guangnan both tacitly agreed on. Ni Guangnan didn't know why Boss Jia brought this up again.
"Teacher, do you remember what we did on the TF card?" Boss Jia suddenly asked with a slightly bitter smile.
"What?" Ni Guangnan really didn't know.
"It's to prevent others from discovering graphene, so we added extra materials!" Boss Jia rubbed his temples.
"Oh, isn't this the idea of that little bastard Liu Nian? I heard that he's torturing those western scholars. They're still desperately researching germanium compounds! You guys are too evil! "Ni Guangnan laughed.
"So, I'm thinking, could ASML have done the same thing to us?" Boss Jia said lightly.
"What, you're saying that the data we got from ASML's machine is fake!!" Ni Guangnan suddenly stood up and looked at Jia Yapeng in disbelief.
"It's not fake. Compared to the real data, it's just a little bit off!" Boss Jia shook his head. "After all, if the difference is too big, it's impossible to fool us."
"That's impossible. If the data from ASML's machine is fake, then how can their lithography machine make a perfect chip?" Ni Guangnan asked in confusion.
"It's simple. Either the data is fake, but the actual manufacturing process is not fake. For example, the data displayed is 44.88667234 degrees, but in fact, when you operate it, the system will automatically increase the degree by about 0.00001 degrees. This final degree is the real correct data, but what you see in the operation panel and database is the wrong degree. "
"Also, the program itself is not fake, but there's a difference of 110000 degrees on the pad. It can't be detected without a high-precision instrument. But who would test a lithography machine like that?" Boss Jia said, "You don't need too much fake data like that. Just four or five of them are enough to completely destroy all the attempts of the imitators."
"But if that's the case, how do you explain that there's nothing wrong with us doing experiments without ASML's machine?" Ni Guangnan asked in disbelief.
"Teacher, other than this experiment, we've never tested these experiments at the 70nm level before. After all, we could only reach the 120nm level at most and verify it according to the 120nm level."
"If that's the case, the precision will be several levels lower. As long as the data is not too different, there won't be a problem! But once the precision is high, it will become a complete disaster. "
"Now that you mention it, I seem to remember …" Ni Guangnan's expression stiffened. "The chips produced by the 120nm lithography machine we produce always have some angles wrong. At the same time, there are more damaged transistors than Western products. The overall product quality is not as good as Western products in any case."
"It's just that I thought there were still many details in the manufacturing process, so it's normal that it's not as good as Western products. I didn't think much about it. I didn't expect that the data used to control the manufacturing process was wrong. The 120nm equipment also uses ASML's 70nm data!" At this point, Ni Guangnan's face was already pale.
"That's right, teacher. I suggest that we immediately mobilize all forces to review and recalculate the data using ASML. At the same time, I'll also send program experts to check the underlying program code of ASML to see if they can find the hidden secrets in it!"
You've already exceeded your reading limit for today. If you want to read more, please log in.
Login
Select text and click 'Report' to let us know about any bad translation.
