novelraw Chapter 20 Computers
Chapter 20 Computers
Having the backing of the state makes things easier.
However, within the system, one must also abide by its rules.
What Lin Yu did last night would be a serious violation of discipline if it happened within the system.
But now Lin Yu is a natural person, and he can do whatever he wants.
It was just a scam.
Even if it is exposed in the future, it will not implicate the country.
I guess he likes playing pranks.
He even bowed and apologized! What more do you want from him?
"Mentor, I'd like to get a computer."
Professor Song didn't care about this at all and simply took out a bunch of keys and tossed them to Lin Yu.
"Go to the R&D facility and get it yourself, just remember to lock the door."
Lin Yu originally wanted to ask Professor Song for some money to buy his own computer, but when his mentor gave him one directly, Lin Yu didn't stand on ceremony.
China Unicom's 2026 AI network will only have one unit, but it still plans to popularize common AI algorithms to the country.
In later years, many large companies open-sourced their code, and Lin Yu also conducted in-depth research on AI for a period of time in order to use it to assist scientific research.
The more I learn about them, the more I realize that the creativity of the Chinese people is truly unparalleled in the world.
Later generations also faced the limitation of having their computing power blocked by the West.
OpenAI controls most of Nvidia's computing hardware and built ChatGPT purely by leveraging computing power.
Despite this significant lead, the United States has invested $5000 billion in the Stargate project.
At that time, China lagged behind the United States by about seven years in terms of the artificial intelligence available on the market.
But the situation can be reversed in an instant.
The company behind Deepseek relies solely on algorithms that are far ahead of their time.
They achieved a remarkable turnaround and took off using only ordinary computing hardware!
After diligently studying the algorithm, Lin Yu realized that compared to the Americans' accumulation of computing power, what real skill was.
In artificial intelligence processing models, computation is the only solution to achieve the goal most of the time.
Whether you ask about using dynamic programming to create a model for managing time among several girlfriends.
Or they asked what Emperor Wu of Han's name was.
These two problems are vastly different in difficulty for people.
However, in artificial intelligence processing algorithms, their calculation paths are similar, all relying on the Transformer model.
Doesn't that sound ridiculous?
In an era when AI assistants were ubiquitous, every inquiry led to a significant waste of resources due to inefficiency.
But Deepseek has offered a new solution.
The algorithm achieves the goal of targeted extraction of target information from a vectorized knowledge base.
The computing power of 1994 did not support achieving the model through computation, but Lin Yu could learn from the design ideas of algorithms to barely create an artificial intelligence that could move, produce, record, and analyze.
Most importantly, this stuff is universally applicable once it's made.
Lin Yu took the key. He hadn't mentioned it to his mentor beforehand, and he hadn't expected his mentor to be so generous.
Since the professor gave him a computer directly, he didn't need to insist on getting one himself.
"Teacher, which computer uses the latest Pentium CPU?"
"The two closest ones to my workstation are both."
"What will you use it for?"
Professor Song then provided another crucial piece of information:
"If you don't need to take it, just leave it there for now. You can go and use it yourself."
"I received news that in April of this year, the government will equip universities with global information networks. At that time, the computers in our group will be able to directly connect with other universities and chat freely on the same communication network."
"If you take the computer away, you won't get this perk."
Was the Internet originally developed and gradually opened up from this node?
That's developed incredibly rapidly.
Lin Yu remembers that back in 04, his elementary school classmates were already talking about going to illegal internet cafes together.
His hometown is very remote, but will the Internet be able to go from initial experiments to every household in just 10 years?
One can only say that one should never doubt a country's strategic vision and its ability to act.
When encountering mountains, they cut through them; when encountering rivers, they build bridges.
Wherever people cry, a brand new road will be paved to their doorstep.
In areas where roads have not yet been paved, judges would carry the national emblem on their backs and ride horses to serve the people.
Lin Yu returned to this era, also wanting to contribute his own spark to this magnificent path of national rise.
Thinking about this, it's certain that if you can still connect to the Internet, you can still connect to the Internet.
He really misses the feeling of chatting freely with others across spatial distances.
Lin Yu then said goodbye to the professor and prepared to type out the code he had envisioned.
But the professor wouldn't have it:
"You're just leaving like that?"
Professor Song slammed Lin Yu's notebook onto the table.
"I've also studied quantum mechanics."
"Come on, explain to me what terahertz is first."
"Teacher, I also have a heart that longs to learn."
Now it was Lin Yu's turn to be dumbfounded. What did he know about terahertz? Before connecting to the knowledge base of the future, he was really just bluffing.
But Lin Yu had no choice. Professor Song simply refused to believe Lin Yu's claim that he didn't know much about the subject and thought Lin Yu was just making excuses.
Finally, after Lin Yu racked his brains and extracted all the knowledge he had left in his mind, Professor Song was satisfied. He then took out two more books on quantum mechanics from the bookshelf, sat back on the small sofa, and waved to Lin Yu that he could leave.
Lin Yu arrived at the research group at 9:30.
While they were discussing terahertz, Professor Song called and ordered a meal to be delivered to the research group for the two of them to eat.
It's already 3 PM now that we've come out.
Time wasn't pressing, but Lin Yu always felt there wasn't enough time, and he was in a hurry to bring all the knowledge back to this era.
As Lin Yu studied the direction of writing the algorithm in his mind, he walked towards the underground research and development site.
The warm afternoon sun was shining brightly, and some of the snow on the roadside had melted.
The system in my mind kept displaying progress prompts.
[55%][66%][78%]
They calculated the possibility of Lin Yu turning code into artificial intelligence.
Until Lin Yu actually arrived at the research and development site.
None of them have progressed beyond 80%.
This is still code based on a parallel CPU scheme. If the target is considered as a single-CPU computer, the success rate of the code drops sharply to below 10%.
However, this doesn't matter to Lin Yu.
He can't wait until the success rate is 100% before writing the code; writing code also takes time. If he only starts writing it then, he doesn't know how many days it would take.
Let's type out a part first, and then we can adjust it gradually later.
After all, it will take him some time to save up enough money to buy a bunch of CPUs. He can't waste this time. He might as well work hard as a programmer since he's idle anyway!
There weren't many people at the research and development site. After the satellite explosion, the supervisor gave everyone a year-end holiday. The Spring Festival was approaching, and no matter how difficult the Spring Festival travel rush was, they still wanted to go home for the New Year.
The empty research and development site was filled with the cold, fishy smell of gasoline and dust.
It was cold, empty, and deathly silent.
But Lin Yu found it quite peaceful, allowing him to focus on coding.
Meanwhile, Yang Jia spent the whole day at the market, begging and pleading until he finally managed to get enough receipts.
Fortunately, Intel's total supply to China in 94 was extremely small, with less than 1000 Pentium 75 chips on the market.
She worked hard and bought 567 pieces.