Desperate Hacker

Chapter 436 Computing Bottleneck

The gene sequencing and start-up companies introduced by Annie Yang are concentrated here in Los Angeles.

So this time Chen Qingfeng no longer has to travel to the east coast. Instead, he enjoys the California sunshine.

Chen Qingfeng lives in a place similar to a resort.

He rented a vacation cabin here, and when he needed to go to the hospital or other places, Luo Chenghui drove him away.

Chen Qingfeng didn't want to hide from peace, but his life alone was a bit boring.

In addition to playing golf, you can also go fishing at the resort.

It's a pity that Chen Qingfeng doesn't like playing golf.

Because the doctor told him that those seemingly green lawns actually use a lot of insecticides, pesticides and the like.

Those things are not good for Chen Qingfeng's health.

As for fishing? Anyway, Chen Qingfeng doesn't eat the fish he catches, he just does it for fun.

Because most of the time he puts the fishing rod on the boat and takes a leisurely look at the computer, looking at code and other things.

Things like code may seem boring to a layman.

But sometimes, you can learn some genius ideas from them.

Of course, Chen Qingfeng has been working over the years and has become somewhat out of touch with his studies.

Today's technological development is no longer what it was 5 years ago.

In fact, Chen Qingfeng feels that even a very smart person cannot keep up with the speed of technological development today.

But after many new technologies come out, the principles used are quite simple.

For example, the convolutional neural network has become popular in the past few years.

The principle used is the compound eye function of animals.

The compound eyes of animals are simulated through mathematical techniques, and then the received signals are recalculated using convolution.

Finally, something like a model is generated, and these training sets are used to achieve the function of identifying some information that can only be recognized by the naked eye.

There is also big data technology.

The earliest application was the search engine company.

Originally, there was a set of algorithms behind the search engine.

Simply put, it is a scoring mechanism for web pages.

But then things got more and more complicated.

So that people have to rethink how to arrange data.

Then there is the connection between research data and data.

Later, when various technologies were combined, big data technology emerged.

Each technology has its own unique application scenarios and realistic application requirements.

It is the needs of real people that drive the development of technology, rather than technology driving real society.

When Chen Qingfeng used to be an engineer, he always had the idea that technology was above all else.

But now, he has realized how naive his original idea was.

In the past few days, Chen Qingfeng has been studying the recommendation algorithm of the British public relations company.

Simply put, the entire program consists of two parts.

The first one is to capture the user's various model information.

Combine this information for analysis.

These users are then divided into categories.

Then there is a user information base. Then the public relations company will accurately recommend something to the people who need recommendation based on the client's requirements.

Although the entire data analysis model sounds simple, Chen Qingfeng had to sigh with emotion. That seemingly inconspicuous consulting company really has something behind it.

Because most on the internet, especially after a while. It will gradually start to become silent.

Simply put, these people will only comment on things that interest them, and they will not post any news on their homepages. Because the majority of people who are not paid attention to are the main force on the Internet.

So every like and every comment from these people basically represents a certain tendency of theirs.

For example, the topic of LGBT and the topic of gun ownership that foreigners are concerned about.

There are also issues involving race and refugees.

Of course there are many more topics.

For example, welfare, medical care, education, employment, and the economic situation!

This information is like weather conditions. It seems difficult to figure out, but with a considerable data sample and powerful computing power, a preference from the group can be summarized.

While Chen Qingfeng was studying this algorithm, he also kept in touch with Susan.

Both felt they had gained a lot from this.

"PEAKER! I just want to say that the person who designed this algorithm is really a genius!"

"I agree! Now we need to implement this algorithm again in our own way!"

"Then this workload is not something I can complete alone!"

"You must have made a lot of money last time by shorting the Adani Group, right? Do you want to open a company of your own and be the boss?"

"I have always had a studio! It's in the Cyberport here in Hong Kong, and it's called Victoria Harbor Data!"

"Oh?"

"Don't worry? I'm not afraid that you will confess to me after being caught by the CIA!"

"You really trust me! Then I guess you have been asking your subordinates to find ways to implement this algorithm?"

"Yes! But what I want to say is that we have encountered some difficulties. We currently have no way to obtain enough computing power..."

This is indeed a trouble. The operation of these algorithms requires a lot of computing power, and nowadays, the improvement of computing power is mainly through GPU.

In fact, there is a once-and-for-all solution, and that is quantum computers.

But that thing is really a bit ethereal.

It is still far from practical application.

In addition to GPUs, some large companies are also independently designing their own computing chips.

For example, Google's TPU actually abandons electrons and uses light to perform calculations. Not only does the energy consumption drop a lot, but the computing performance also improves dramatically.

In fact, the hacker group originally referred to technical lunatics who had the courage to make breakthroughs, rather than people who made trouble everywhere.

But then there were more hackers doing bad things, and hackers turned into people who caused damage.

So people created a new word called geek.

In the early years, Facebook's data center was built by some hardware geeks without any precedent.

Later, a series of technologies were developed that gradually made data centers popular.

"PEAKER! We can't be trapped by algorithm problems forever? I wonder what you can do?"

"Susan! Have you heard of the data center that Microsoft threw into the ocean?"

"I've heard of it! But I don't know what kind of technology they used?"

"Ask that guy from Duke. He works on Wall Street and should know something! At least we need to know what technologies we need to acquire to meet our needs for computing power and data centers?"

……

Chen Qingfeng quickly sent the message to Duke, but the guy replied:

"I have to look up this kind of thing before I know it. Don't worry, I will help you find out!"

(End of this chapter)