Technological power starts with upgrading lenses

When it comes to chips, the first thing most people think of is the CPU processor.

As the brother of the CPU, the graphics card GPU is often ignored.

It is probably because the CPU can also do the work that the GPU does, so except for gamers, no one pays much attention to the performance of the GPU.

Especially when I saw that the graphics card was more expensive than the CPU, I didn’t even install a discrete graphics card.

But when your requirements for image processor computing power become higher and higher, the importance of GPU becomes apparent.

In addition to games, GPUs play a pivotal role in fields such as imaging, artificial intelligence, and even aerospace.

In terms of division of labor and composition structure, the CPU is like the designer and supervisor of a house, while the GPU is the hard worker.

Designers and laborers have one thing in common - they are both human beings, with hands, feet and brains, but they are just good at different things.

Because the GPU is mainly composed of ALU transistor units, the high degree of repetition makes it less difficult to design and manufacture than the CPU.

But the reality is that graphics cards are more expensive than CPUs.

Today, Invida, the world's major graphics card manufacturer, has a market value of US$800 billion, four times that of Intel...

This is magical.

Mainly because Weida is so lucky.

Under the two super trends of virtual currency and artificial intelligence, Invida was suddenly promoted to the sky.

As for why those big manufacturers that can make CPUs don’t make GPUs?

Just like car companies that manufacture fuel vehicles cannot suddenly build new energy vehicles, major CPU manufacturers such as Intel also need time to react, but by the time they react, Invida has already secured half of its market share.

This is true for Intel, let alone domestic chip manufacturing. Currently, those engaged in graphics cards are basically assembly foundries.

Nowadays, these two trends are still there, public and private, and Kang Chi also wants to make some achievements in this area.

From a longer-term perspective, Kangchi will definitely enter the manufacturing fields of cameras, mobile phones, computers, etc., and these cannot avoid chip manufacturing. If we plan ahead now, we will not be so passive by then.

Therefore, his current goal is to first create a graphics card that is easiest for him to create, and then use common experience to upgrade this prototype until the level is almost improved. Then, through process analysis, he can obtain the entire industry chain. Technology.

By constantly reviewing domestic and foreign materials and papers, Kangchi quickly found relevant information on 2D accelerator cards.

The design of this outdated thing was not too complicated. After pondering it carefully for a few hours, he had a vague understanding of its principles.

However, the driver algorithm gave Kangchi a headache.

[To determine which of two pixels is closer to the path, you need to test the difference in the offsets of the two pixels:

Dlower-Dupper=2m（Xk+1）-2Yk+2b-1

…]

Looking at the paper that looked like a book from heaven, Kang Chi felt like a primary school student.

Even if he finds a usable algorithm, he probably doesn’t even know about it...

"Forget it, once the graphics card is built, just find someone to write a driver."

Professional matters must be left to professionals.

The next step is to consider how to actually build a graphics card.

Kang Chi checked the information again, and after understanding the chip manufacturing process, his head became even bigger.

too difficult……

From manufacturing monocrystalline silicon wafers to photolithography, etching, and packaging, no link in between is simple!

Where can I find equipment that can make chips?

Kang Chi thought hard and suddenly remembered that when he was in college in his previous life, there was a school-enterprise cooperation project, and those cooperative companies could use the school laboratories.

So with the attitude of giving it a try, Kang Chi sent an email to the Microelectronics Research Institute of Suzhou University, which is relatively close to Dongyang and is very professional in semiconductors, to see if he could get some help.

——

The next morning, after Kang Chi and Fang Qiming made simple preparations, Daqin Optics' lens production line began a trial run.

After being trained by Kangchi, Fang Qiming quickly learned to control these three large machines.

The division of labor between the two of them is very efficient, with Fang Qiming manufacturing parts and testing, and Kangchi responsible for assembly.

But problems soon arose,

The production capacity of these three machines is 40 shots per hour,

However, the assembly speed of Kangchi can only barely reach 10 pieces, which is completely unable to keep up with the speed of the machine.

Fang Qiming, who was only responsible for delivering information, had a lot of free time. Seeing Kang Chi busy, he couldn't help but suggest:

"Brother Kang, how about you teach me how to assemble it? The two of us will assemble it together. Once we get familiar with it, we will definitely be able to keep up with the speed of the machine."

Kang Chi stopped what he was doing and thought for a moment.

He first made a calculation silently in his mind.

Calculated based on 8 hours of working time per day, this production line can produce 320 lenses. If it works 22 days a month, it will be 7040.

If all these lenses are sold, the turnover will be around 600,000 yuan.

In terms of costs, miscellaneous items such as raw materials, rent, water and electricity, and express delivery fees account for about 50%.

The store policy of free replacement may not have much impact in the early stage, but in the middle and late stages, it may be a big burden.

After the recycled lenses are converted to general-purpose experience, some materials can be put into the machine for reuse, which can save a certain amount of cost, but some materials are waste that cannot be recycled.

Therefore, it is estimated that 10% of the money will be spent to subsidize the free replacement activity.

So the remaining 40% of the profit, about 240,000, is the funds he can use, for example, to recruit people.

If you think about it this way, it's quite generous.

Of course, the premise is that all the goods produced can be sold.

It’s just that Kangchi is very confident in this!

Will such a high-quality and low-priced product fail to sell?

After settling the accounts, Kang Chi said to Fang Qiming: "We can definitely recruit people to do this kind of highly repetitive work. You only need to be responsible for managing the factory and store in the future."

It’s finally time to recruit troops!

Fang Qiming immediately became excited after hearing this.

This also means that he is on the verge of promotion!

Here Kang Chi just did what he said. He directly opened his mobile phone and contacted HR on the recruitment website and put forward his needs.

"A worker responsible for transporting materials, our materials are not very heavy, can earn a monthly salary of 6K;"

"The requirements for four technicians who assemble products are not too high. They only need to be able to repair digital products such as mobile phones. The monthly salary can be 8K,"

"Finally, I want someone who is responsible for product quality inspection and delivery, with a monthly salary of 8K. We work 9 to 5, with weekends off and social security. Food and accommodation are not included."

The HR here was immediately shocked after hearing the requirements and treatment.

"With such a high salary, does it also include weekend breaks and social security?"

"Well, is there a problem?"

"No, I'm just a little surprised. You should know that most factories in Dongyang don't buy social security for their employees, let alone any weekend breaks..."

"Well, don't worry about this. We will do what we say and sign the contract! Oh, by the way, we are a technology company and we need to sign an additional confidentiality agreement."

"clear!"

Due to the good salary and good benefits, the HR of the website quickly found a large number of suitable candidates and arranged interviews for Kang Chi the next day.

After a round of interviews, Kang Chi directly finalized the candidates, and then they started training and on-the-job training.

In less than a week, the entire product line could be produced at full capacity without Kangchi.

At the same time, the product details page of the second-generation lens in Daqin Optics’ Dabao store also quietly changed.

The first is to cancel the purchase limit, and the second is to update the new product introduction, using a line of banner posters to read:

[The second-generation products are completely refreshed, with improved texture and unchanged price! ]

Well, I saw some readers commented that the names of the protagonist and the lens brand were a bit strange, so I changed them to homophones, hoping it would look better~ (I have difficulty naming, I used an assistant to randomly name them, so I’ll try to be more careful in the future. )