My fintech empire.

Chapter 1235 [Computing Resources Are the Key]

At this moment, Fang Hong was not stingy with boasting, because it really exceeded his expectations, and exceeded his expectations by a large margin.

This three-dimensional chat room can be said to have brought people from all over the world closer together, shortening their sense of distance, participation, and immersive experience, greatly enhancing their sense of involvement.

Fang Hong asked: "Is this the only scenario in this kind of chat room at present?"

Lou Gaoyuan replied: "There are already more than a dozen scenes, including indoor, outdoor, space backgrounds, etc. More scenes will be developed later."

After saying that, Lou Gaoyuan immediately changed the scene. This time, it took about nine seconds to load. When the new chat room scene appeared, several people present were slightly surprised. Among them, Hou Junda, who was in charge of Yixing Video, subconsciously exclaimed: "Fuck!"

Hou Junda was shocked and immediately complained: "I said, Lao Gao, could you please give me a heads up so that I can be mentally prepared?"

The new scene switched to a panoramic outer space background. Everyone was sitting on a single lazy sofa, suspended in the space. Fang Hong looked around the scene, with the deep starry sky under his feet. No wonder Hou Junda was startled by the suddenness.

At this moment, they seemed to be suspended in space, and high in the earth's orbit. The earth was on the side, and it was slowly rotating visibly to the naked eye. At this time, an artificial satellite appeared and passed by the background next to them along the orbit, gradually moving away.

Fang Hong praised repeatedly: "Well, it's very good, it has a bit of the flavor of a local virtual world."

At this time, Chen Yu said: "The development and operation costs are very high. These immersive scenes are all made with real money, and they are only single local scenes. This is the limit of what we can do at present. The real virtual world cannot be supported by the existing computing resources."

Lou Gaoyuan agreed: "Chen Yu is right. The cost is very high and the operating cost has also soared."

Chen Yu also continued: "My second-generation large model now has a daily operating cost of more than 200 million yuan, which is a fixed expense. The daily operating cost of the third-generation large model has soared to 10 times, and this is just playing with the text language generation formula. By the fifth-generation large model's video generation formula, the daily operating cost will definitely soar to million or even million."

You have to know that this is one day's operating cost, not one month, let alone one year. If the daily operating cost of the fifth-generation model is calculated at 3 million yuan, the operating cost alone will be more than billion yuan in a year.

At this time, Qin Feng said: "A major pain point now is the severe shortage of computing resources."

Chen Yu nodded and said, "Yes, the main obstacle to the further development of artificial intelligence, including the establishment of the virtual world industry chain, is the lack of computing resources. We need more powerful computing chips, especially since the operation of the virtual world requires too much computing power. It is simply a computing power black hole."

"Although our current scene is amazing in the contemporary world, it is not a real virtual world strictly speaking. It has to be preceded by the word 'pseudo', which means it is just a pseudo-virtual world. To achieve a truly immersive experience in a virtual world, it requires not only technological breakthroughs in brain-computer terminals, but also huge computing power requirements. It would be difficult to increase the computing power resources by 10 million times on the basis of the current computing power resources."

After a pause, Chen Yu added: "In general, under the existing computing architecture, it is almost impossible to achieve the huge computing power required to achieve the ideal state."

Finally, Fang Hong said, "This means that such a huge demand for computing power can only be solved by breakthroughs in the underlying technology of quantum computers, right?"

Chen Yu nodded. "It can be said that the existing computing system, whether in terms of energy consumption or efficiency, is almost overwhelmed by the huge computing power demand. Without a breakthrough in the underlying technology, silicon-based chips and von Neumann architecture are basically the ceiling of artificial intelligence, and the power consumption is also particularly terrifying, which is like using a horse to pull a high-speed rail carriage."

Obviously, Chen Yu’s view is very clear. If we want to realize a true virtual world and strong artificial intelligence or general artificial intelligence, we still need breakthroughs in underlying technology. The computing power of classical computers under the von Neumann architecture is limited, and the consumption of energy and electricity is also exaggerated.

The underlying technological breakthrough he mentioned is quantum computers. Matrix Quantum is also working hard on this field. The company's R&D investment last year was second only to Xingyu Technology, which is also an exaggerated scale of investment.

Finally, Fang Hong said calmly: "It's still too far away to talk about the virtual world, strong artificial intelligence, and quantum computers. I have confidence in you, but it's unlikely that there will be a major breakthrough in a short period of time. I know this, and you know it even better."

Everyone nodded.

Fang Hong continued, "Don't be in a hurry. Take your time. We can slowly study quantum computing and strong artificial intelligence. Now we can go to a 'pseudo-virtual world' which is also a huge leap. We are still ahead of the world and others by a huge margin. The key is to prove whether the current path is feasible. What is the biggest gain now?"

"It was Chen Yu who proved that this approach is feasible. For us, the remaining issue is engineering. We are not yet leading the world in terms of overall computing power, but Matrix Quantum has not stood still in recent years. Although the computing power chip is not leading the world, it is still useful. If the computing power performance is not enough, we can make up for it with large quantity."

"We Chinese have a very powerful characteristic. Industrialization and engineering are our absolute strengths. No one on the surface of Blue Star can compare with us. Computing chips are also infrastructure after all. We are responsible for infrastructure construction. Under the existing technical architecture, we are working hard to build production lines, increase production capacity to produce more chips, make computing boards, assemble servers, and build a computing center. This is an industrial process."

"As long as the path is proven to work, it's fine. Industrialization and engineering are definitely not our opponents. We should regard it as an engineering construction problem. Whoever lays down more 'rebars and cement' will win. The question of responsibility is also simple and direct. We should use our infrastructure maniac attributes to build data centers and computing centers. If one computing center or data center is not enough, then we can build two, three, five, ten, or even a hundred. This is what we call flying bricks with great force!"

"In addition to computing power, chips and memory, electricity is also a big problem. Laomei can't handle the electricity cost of such a massive server cluster. The cost is also higher than ours. They can't afford such expenses. In terms of power generation and electricity costs, we are definitely superior. In terms of electricity, Laomei can do what we can do with 1 yuan."

"Of course, what I said is based on the existing chips and von Neumann architecture computing system. It can only be regarded as a transitional solution, not the final solution. The final solution still depends on when Chen Yu's quantum computer research team lights up this technology tree, as well as Qin Feng's brain-computer terminal. As for the virtual world, such high-end technology is not needed yet."

When everyone heard what the big boss said, they basically understood what he meant. That is to do it, even the pseudo-virtual world must be done. This serves as a link between the previous and the next.

……

(End of this chapter)