Back in 1978, I was admitted to Northwestern Polytechnical University.

Chapter 272 Huaxin Company, the Pioneer and Monopolist of GPUs

The first generation of CRT technology was a spherical screen, the second generation was a flat right-angle screen, and the third generation, or the ultimate version, is a flat-screen monitor!
So-called flat-screen monitors are not actually truly flat; they still have a slight curvature. Therefore, in the late 1990s, the third generation of flat-screen monitors appeared. These monitors are perfectly straight in both the horizontal and vertical directions, with no curvature visible. Image distortion and screen glare are also minimized.

LG's Future Window and Samsung's Dyna technology are representative examples of flat-screen monitors.

However, these are not the best either.

In fact, the reason for image distortion is quite simple. After the electron beam is emitted, it passes through deflection coils and begins to scan the screen left and right and up and down. Due to the different distances from the screen, distortion will occur at the edges.

The solution is to completely change the structure of this deflection coil!

The circular deflection coil is replaced by two rectangular deflection coils, separated horizontally and vertically. Simultaneously, under the action of the square pyramidal glass shell and the triple plane, the deflection of the electron beam is greatly corrected. It can be said that every point on the display can be accurately focused.

In this way, it becomes completely flat; this is the technology behind square tube displays.

By this time, CRT monitors had finally reached their limit. Unfortunately, it was too late. By the time large-sized cathode ray tubes had reached their peak, LCD technology had matured! As a result, bulky monitors were quickly phased out, and both televisions and computers adopted flat-panel displays. These were truly flat, as they were composed of individual light-emitting dots.

Now, in the 1980s, we'll have developed a flat-screen monitor. By then, televisions from the East will become the most popular product in the world! We'll register the patent, and other companies that want to produce them will have to pay us patent fees!

Previously, a key issue was that the weak production technology in the East could not solve the problem of large-scale industrial production. Now, with the production lines exported by Thomson Company, this production problem can be solved!
Regarding flat-screen displays, Qin Liang didn't introduce the specific technology, but only mentioned the advantages of this technology, which excited everyone present.

"Qin Liang, Northwestern Polytechnical University is truly amazing! If you ever need to teach us this technology, we're willing to pay for it..."

Qin Liang nodded: "Of course, the technology can be licensed to you, or to other domestic television manufacturers. Director Wang, as I said before, a monopoly is not good. Only when a hundred flowers bloom can our country's development be promoted."

In those days, everyone was proud to buy imported color TVs, but in later generations, who still buys imported TVs? Perhaps some do, but they are extremely rare! Because our Eastern color TV industry has flourished!
This technology will be handed over to the corresponding department at Northwestern Polytechnical University for research and development. Once it is developed and patented, it must be registered under Northwestern Polytechnical University's name. The domestic licensing fee should be symbolic, but it absolutely cannot be given to Changhong alone, allowing them to dominate the market. That is not what Qin Liang wants to see.

That's why, just now, when Changhong Factory wanted to give Northwestern Polytechnical University shares, Qin Liang refused. Once there's a conflict of interest, things will lean towards Changhong Factory in the future.

"Liangzi, tell me, what benefits do these things bring to our army?" Old Lin asked.

You've done all that work, and in the end the Air Force doesn't get any benefits at all...

"Old Lin, there are so many benefits to this. Our Air Force's modernization and informatization definitely require computers and monitors. With this 21-inch monitor, things will be much more convenient. It will also be much easier for us to do computer-aided design with a large screen..."

This large monitor is really necessary! Although Qin Liang has gotten used to the current spherical monitors, he still misses the 32-inch 4K screens of the future. Although he can't handle LCD monitors yet, making a flat screen would be convenient for him.

Yes, I really do have that selfish motive...

After speaking, Qin Liang looked at everyone present and said, "Everyone, you must remember your promise. Once the production line is in place, we will produce a batch of cockpit displays for the Air Force."

"Of course, we haven't forgotten, Comrade Qin Liang, you can rest assured."

Why would Qin Liang help Changhong out of the blue? It was for the Air Force's cockpit display device!
Although the JF-17 Thunder, which is now making its maiden flight, has a glass cockpit, it is produced using domestic technology and its performance is not good. While Thomson's technology may not be as advanced as that of the island nation, it is still far superior to that of China.

Changhong factory must make a statement!
"Comrade Qin Liang, regarding the integrated circuits needed for this television set..."

"Don't worry, once our Zhenhua Electronics' new production line is in place and has sufficient capacity, we can start production. Other factories in China can also do it, such as the Shanghai Components Factory No. 5..."

Last time in the capital, when Huaxi Company was established, Shanghai Components Factory No. 5 also showed its support. Qin Liang didn't mind giving a word of advice to such a sensible company: their attempt to copy the 8086 processor was not going to work. They might as well copy the chips needed for televisions, which were much easier to make than processors.

"Okay, thank you, Comrade Qin Liang."

After making a trip to Chengdu, Qin Liang returned to Chang'an in mid-April, where the surrounding area was bursting with the vibrant colors of spring.

Back at Northwestern Polytechnical University in high spirits, Qin Liang received surprising news: Yu Youcheng had brought back more than fifty Chinese engineers from Intel!
The mathematics department allocated an entire floor of the teaching building specifically for them! They were also given priority in being allocated housing in the newly built faculty housing building. If they had family members or children who came with them, their work, study, and living needs would be taken care of.

Director Meng of the Mathematics Department was grinning from ear to ear; he'd probably be laughing in his dreams.

Ni Gong has already gathered all the chip experts in China, and Yu Youcheng has also brought over experts from Intel. In the future, Yixi will become an even more dazzling presence and a world-class chip design and manufacturing organization!

When Qin Liang entered, everyone was gathered together. Director Meng was holding a welcome party for everyone. Upon seeing Qin Liang, he immediately said, "Everyone, upon learning of your arrival, Comrade Qin Liang rushed back from out of town!"

Swish, swish, swish!

All eyes turned to them.

Qin Liang could only smile as he walked in: "You are all experts in the chip field. You came to Dongfang to build advanced display chips. Now, I have some ideas. You can discuss them later and see if I am right. First of all, I think that display chips and central processing unit chips need to have completely different internal architectures because they need to process different objects."

Everyone became serious. What Director Meng had just said was all nonsense. They were here to research advanced technology! What Qin Liang was saying was much more to their liking!

Although there are no dedicated graphics cards yet, computers still need to display content, so similar hardware does exist.

Leaving aside the earlier ones, in 1978, Motorola launched the MC6845 video address generator, which was then used in IBM and Apple personal computers. In 1981, Motorola developed the MC6847 video display generator. Computers of that era were almost all using these.

Similarly, Commodore's MOS Tech subsidiary is also using the same solution to provide graphics card output for home computers.

However, these are all monochrome, and at most they have some built-in functions for generating circles, lines, etc. It can be said that they are still very primitive.

As for how to design advanced 3D display chips, they had no idea. Now, Qin Liang went straight to the technical topic, and everyone was very excited.

"To achieve 3D calculations, it means processing a large number of triangles... Am I right?"

Those complex 3D graphics are actually made up of small triangles, which belongs to the fields of art or mathematics. Qin Liang is not a professional in this field; he has only heard of it.

Everyone nodded: "That's right."

"I think the current CPU architecture is not suitable for performing a large number of triangle calculations. We can design a very small unit that is responsible for generating triangles, and then integrate a large number of units on a chip so that they can work together to meet the requirements."

This is the difference between a CPU and a GPU. The former needs to process various types of data and perform complex logical deductions, so the CPU core is very complex, and in this era, it still typically has only one core. On the other hand, the data that a GPU needs to calculate is relatively simple, but the amount of data is enormous. While a CPU can indeed perform the calculations, its efficiency is very low. In contrast, a GPU, with its numerous cores, can perform calculations at lightning speed.

Everyone's eyes lit up.

"Qin, do you mean that the display core uses parallel computing?" Yu Youcheng asked.

"That's right, it's parallel computing! We can continuously integrate more cores as our chip manufacturing process advances. Of course, the triangle is just an example; in reality, it will definitely be more complex. In short, this is my general idea. Another thing is to invent dedicated graphics display functions and reserve interfaces for connecting to computers..."

With the hardware in place, the next step was software. In the 1980s, there was a plethora of graphics card manufacturers, but in later years, the number gradually decreased because many technologies were proprietary and many programs were monopolized by a few large manufacturers.

For example, OpenGL, DirectX, etc.

Other manufacturers are out of luck. Even if they invent it, without supporting software, it's useless. For example, take Huawei phones in the future. They create a pure HarmonyOS, but without any compatible software, how can they use it? It can only be used as a backup device.

Now, if we at Northwestern Polytechnical University were to work on these things and strive for a monopoly in display technology, we would definitely be more successful than Huang in later generations!
Moreover, the development of this technology has very practical significance. It's not just about making games. In the future, fighter jets will have transparent cockpits, allowing people to see right through the fuselage by looking down. This will rely on advanced video processing technology. Even if the cockpit displays need to show complex 3D maps, it will still depend on this technology, not to mention the artificial intelligence that will become so popular in the future...

Qin Liang explained his ideas at length, while Director Meng and the others could only watch. The engineers who had returned from abroad all stared wide-eyed, and some were even taking notes rapidly...

Northwestern Polytechnical University (NPU) has plenty of money. With so many successful projects, it has accumulated a lot of R&D funds. NPU has successful models, and Loongson is a typical example. NPU also has a complete technical roadmap. Therefore, display chip technology will definitely be able to develop!
"Qin, we'll develop this display chip according to your design!" Yu Youcheng said. "Next, what name do you plan to give our project? Will you still name it after Loongson?"

Qin Liang shook his head: "No, the name of the Loongson central processing unit is used for graphics processing. Of course, it needs a new name. Do you have any ideas?"

Qin Liang's naming style is simple and straightforward. If it were up to him, he would just pick one from the future: Moore Threads, Muxi Technology, Suiyuan Technology, Shanghai Biren, and Jingjia Micro... As high-end computing power was banned by the Americans, our East also rose to the challenge and independently developed many GPUs.

The earliest of these companies was Jingjia Micro, which began systematically developing GPUs in 2006. At that time, it coincided with the upgrade of the aircraft cockpit display system in Dongfang. Previously, DSP and FPGA graphics accelerators were used, but GPUs were to be used. Moreover, these could not be purchased from outside and had to be produced in-house. Jingjia Micro, which was little known to the outside world, was actually a leader in the field of aircraft graphics display.

The other companies were also quite commendable...

"How about calling it Huaxin?" Yu Youcheng suggested.

The CPU is Loongson, and the GPU is Huaxin; that's a good match.

“Sure!” Qin Liang readily agreed. “Today is the founding day of our Huaxin GPU. Tonight, let’s have a banquet in the cafeteria…”

"No need," Yu Youcheng said. "Now that we have the technical roadmap, let's start working right now. We can hold a celebration banquet for us after our Huaxin No. 1 is developed!"

"it is good!"

People who come from Intel are very dedicated. They don't think about 9 to 5, they only think about 996 or even 007!

Under their guidance, the computer science students assigned to them also threw themselves into the vigorous research and development of the project, and the world's first 3D display GPU began to be conceived from that day on!

Seeing everyone engrossed in their work, Qin Liang walked away with a light step. Director Meng followed behind him, and after walking a short distance, he spoke up: "Liangzi, do you think there might be some among them...?"

"It's really hard to say, so we need to ask the relevant departments to keep an eye on them and their families. Of course, since the entire project is in the computer room, they can try to steal some concepts, but it won't be so easy for them to steal the specific technology."

Once bitten, twice shy. Among these people, there might be Intel's corporate spies, but... we've been prepared for this for a long time, and we can even ask relevant departments to keep an eye on things. Stealing technology isn't that easy!
Although there are risks involved, we cannot shut people out because of those risks; we must be open.

The two talked as they walked and soon arrived at the office area, and then... a hearty laugh could be heard.

"Loongson-1 is the pride of the East! Such a good product cannot be used only on the Air Force's aircraft; it must also be installed on the Navy's warships! We cannot afford to buy an entire Type 42 destroyer, so we can only import some subsystems to equip our Type 051 destroyers and form a dedicated air defense destroyer. This integration work is quite extensive. Among them, we hope to develop the combat command system ourselves, and the processor has already been decided as Loongson!"

Qin Liang frowned instantly. He pushed open the door and saw some military officers in the conference room, accompanied by some teachers from the Computer Science Department, including Engineer Ni.

One of the elderly men looked very familiar.

Just then, the other person looked up at the door and met Qin Liang's gaze. He immediately smiled and said, "Liangzi, you're back?"

"Old Liu? When did you arrive?"

"About ten minutes ago, I heard you were having a meeting with a group of engineers from the United States, so I didn't disturb you. I wanted to find out more about the Loongson R&D team first," Old Liu said with a smile. "Now, a large number of Chinese are returning home to support our country's economic construction and technological development. It's truly gratifying. It reminds me of when the country was first founded..."

Shortly after the founding of the People's Republic of China, the first wave of overseas Chinese returned to the country. They gave up high-paying jobs abroad and returned to their motherland to build it with their own hands.

Now that the country has opened its doors, countless people want to go abroad, while some are going against the tide and choosing to return home. Professor Liu is deeply moved by these people. Northwestern Polytechnical University is a golden brand, attracting the world's brightest minds and accelerating the country's technological development...

Qin Liang's mind was racing.

Sure enough, just like in history, the navy started working on the Type 051S missile. After some negotiations, it was decided to purchase the British Sea Dart air defense missile system to give the navy area air defense capabilities...

The navy is in a tough spot. Almost all of its warships rely on anti-aircraft guns or even anti-aircraft machine guns for air defense, which is outdated. Therefore, the navy needs advanced air defense missiles, but domestic technology is not up to par. The only mature missile is the HQ-2, which requires toxic fuel to be loaded before use and can only engage high-altitude, high-speed targets.

Therefore, purchasing from outside sources became the inevitable choice.

Unfortunately, the Navy chose the wrong target; the Sea Dart was a complete disaster!
Now that I've run into this situation, I have to sabotage it! What should I do?

"Liangzi? Comrade Qin Liang?"

Beside him, Director Meng quickly tugged at Qin Liang, pulling him out of his reverie. He was a little embarrassed: "Excuse me, leaders, I was just thinking, since you're all here, would you like to take a look at our school-run factory?"

Is it a computer manufacturing factory?

"No, they produce missiles."

　PS: The difference between CPU and GPU is that the green ones are individual core processing units.

　　
　
(End of this chapter)