In the morning, visit the supercomputing center.

In the afternoon, Academician Guan's office!

"How is our supercomputing center?" Academician Guan said.

"Very good." Ye Fei nodded and said: "It is worthy of being one of the eight supercomputing centers, whether it is hardware or talent, it is impeccable." ”

"We're nothing." Academician Guan shook his head and smiled: "The best supercomputing center in Xia Country is in Liangxi, if you have seen it, you will be even more amazed." ”

"Hmm!" Ye Fei nodded.

The divine power of Liangxi. The light of Taihu Lake is the fastest supercomputer in Xia and the sixth fastest supercomputer in the world.

Naturally, the Liangxi Supercomputing Center is the most luxurious supercomputing center in the Xia Kingdom.

"In the afternoon, call everyone at the 111 base to have a meeting as your official meeting with everyone." Professor Kwan said.

"Let's not meet yet." Ye Fei shook his head and said: "I want to take a look at the information of supercomputing algorithms first and determine the research direction of supercomputing mathematical algorithms in the future." ”

"Let's study this together!" Academician Guan said.

"No, I'll do it alone."

The function of the system is to obtain knowledge and look at materials by doing problems, and find the knowledge you need from many materials.

Therefore, looking at supercomputing data and finding research directions from them, he can do it alone.

"But there's a lot of information, and you need to look at it for a long time." Professor Guan hesitated: "Are you alone?" ”

"I can!" Ye Feidao.

"Okay, I'll let Xiao Yang take you to the reference room."

"Okay."

After that, Ye Fei followed Assistant Yang to the data room.

It's more like a data room, but it's more like a small library, which is divided into categories for GPU, CPU, and processor.

Of course, this is the big category, and there are many small categories within the big category.

Initially, there are tens of thousands of volumes of information.

Assistant Yang said: "Ye Fei, all the information is here." ”

"I'm sorry for you, Assistant Yang."

"It should."

Assistant Yang left after speaking.

Ye Fei took a deep breath, he didn't need to read all these materials, only the algorithm part.

But even then, there are hundreds of volumes of material.

"It's good that there is a system, otherwise it would have taken a long time to read it."

He doesn't need to look at the information carefully, as long as he glances at it, the system automatically distinguishes the content he needs.

And what is needed, he needs to look closely.

With that, he began to look for the information he needed from the many sources.

He first found the algorithm data, and he glanced at the hundreds of volumes of data.

Pause occasionally to peruse what you're prompted for.

You can't see a minute on each page, you can't see an hour and a half on each volume, and you can read ten volumes in a day.

But his behavior seems strange to many people.

You can read a novel with ten lines at a glance, but look at information at a glance with ten lines?

Are you sure you're looking at the data, not scanning the data?

The people who came into the data room to look for information were very puzzled when they saw Ye Fei's strange behavior.

Academician Guan's office!

"Dean!" Assistant Yang came in and said, "Ye Fei is scanning the data in the data room. ”

In the Supercomputing Center, Academician Guan is the dean, and Academician Li is the vice president.

"What saccade data?" Academician Guan said suspiciously.

Assistant Yang said: "He reads the information very quickly, ten lines at a glance, turns a page after a few glances, and occasionally pauses to read carefully. ”

"Leave him alone." Academician Guan said: "He may have some special reading skills, besides, with the results he has made, I believe he is not such a boring person." ”

"Good!"

March 11th!

Ye Fei finished looking at the data, spent two days sorting out and summarizing the data, and determined the next research direction of supercomputing mathematical algorithms.

March 14th!

Ye Fei found Academician Guan and said, "Dean, you can hold a meeting, and I will talk about the future research direction of supercomputing mathematical algorithms." ”

"Good!"

In the afternoon, at half past one, the auditorium of the supercomputing center!

Thousands of people gathered together, more than 500 people were from the 111 base, and the remaining 500 people were from other bases to see the excitement.

Many people in the supercomputing center are full of curiosity about Ye Fei, and they heard that today Ye Fei will talk about the research direction of supercomputing mathematical algorithms in the future, and they all want to know what Ye Fei will say at today's conference.

Many professors and two academicians from the supercomputing center came.

At this time, the conference had not yet begun, and the people sitting in the auditorium were talking.

There was no host present, and at two o'clock, Ye Fei walked to the podium on the stage with a notebook and connected the notebook to the projector.

"Hello everyone, I'm Ye Fei!"

"Papapa......" There was warm applause below.

Since Ye Fei created mathematical algorithms last year, they have heard a lot of news about Ye Fei.

After that, Ye Fei came to the supercomputing center and solved the problem that had not been solved by the 40 team for three months.

At the supercomputing center, they respect the able.

If you have the ability, you will be respected by everyone, and if you don't have the ability, you don't want to listen to you.

Naturally, Ye Fei is capable in the eyes of most people.

As for how much ability they have, there is no limit in their hearts.

But at least now it seems that Ye Fei is capable.

Therefore, they welcomed Ye Fei to the National Supercomputing Center.

Ye Feidao: "As a student who has been studying at Zhonghu University for six years, I have always been well-known and yearning for the National Supercomputer Dune Center. ”

"I'm excited to be here at the Supercomputing Center today."

"I am also honored that the Supercomputing Center and I have cooperated to create supercomputing mathematical algorithms."

"Next, let me talk about the research direction of supercomputing mathematical algorithms."

"Everybody, look at the projector!"

"The next step in the study is a three-step process."

"First, create a highly compatible mathematical algorithm, and it is a mathematical algorithm that adapts to supercomputing."

"Second, the five algorithms are compatible with mathematical algorithms."

"Three or five algorithms are compatible with each other, and after compatibility, they are compatible with mathematical algorithms."

"The five algorithms are, clustering algorithm, ant colony algorithm, massive algorithm, fish swarm algorithm and particle swarm algorithm."

Clustering algorithm, also known as group analysis, as the saying goes, things are clustered by like, people are grouped, clustering algorithm has an important role in data capture and analysis.

The ant colony algorithm is an algorithm that finds the path to optimization.

Massive algorithms are a kind of algorithm for finding positioning, and many Internet companies will use them.

For example, Baidu's IP positioning, even tens of billions of IPs can be located at the same time.

The fish algorithm is created based on the characteristics of the fish, which tend to go to places where aquatic plants are abundant, and the fish algorithm can automatically find the optimal path.

In terms of characteristics, it is similar to the ant colony algorithm.

But in fact, it is different, the ant colony algorithm has a large range, cannot automatically find the optimal path, and the path finding time is long.

However, the fish school algorithm has a small range and can automatically find the optimal path, and the path finding time is short.

Particle swarm optimization algorithm, also known as particle swarm optimization algorithm, particle optimization algorithm, and particle optimization algorithm.

It is an algorithm created to simulate bird foraging, and it is a random search algorithm based on group cooperation.

Wide range, but random.

Each of the five algorithms has its own advantages and disadvantages, and Ye Fei makes the five algorithms compatible, which can make the five algorithms learn from each other's strengths and weaknesses and become the most perfect algorithm.

"Woo......" everyone exclaimed, this is a bit difficult!

Algorithm compatibility is not uncommon, and the current algorithm of the supercomputing center is compatible with multiple algorithms.

But the mathematical algorithm is different from all the algorithms in the world, it uses a lot of mathematical knowledge and very little computer knowledge.

This limits the compatibility of mathematical algorithms.

The five algorithms are compatible with mathematical algorithms and require a large number of packages in order to be compatible.

If you want to make the five algorithms compatible with mathematical algorithms, unless the underlying architecture of mathematical algorithms is rebuilt and a lot of computer code is added.

But in this case, is this still a mathematical algorithm?

The merits of mathematical algorithms are gone.

Ye Feidao: "Everyone knows that mathematical algorithms are created by mathematics, and the integration and application of mathematical formulas, I think everyone has seen it." ”

For example, 1+1x2, if 1+1 is one algorithm and 1x2 is another, then the two algorithms are compatible applications.

"Theoretically, mathematical algorithms are also highly compatible, and they can be compatible with any algorithm."

"I am confident that the five algorithms and mathematical algorithms will be compatible without losing the advantages of mathematical algorithms."

"In the next work, I will organize a group to create highly compatible mathematical algorithms, and the rest of the groups, I will divide them into five groups, each group is responsible for one algorithm."

"At the beginning, we made sure that all five algorithms were compatible with mathematical algorithms, and then slowly tried to make the five algorithms compatible in pairs, threes, and threes, and finally made them five and five compatible."

"Finally, make the five-five compatible algorithm compatible with the mathematical algorithm."

"I'm done!"

"Papapa......" There was warm applause below.

Although it sounds like a difficult study, which one is not difficult?

A study that is not difficult is not a study, it can only be regarded as a family.