Chapter 304 You May Need a Quantum Computer! (Additional Chapter for the Leader Gao Shan Liu Shui)
With the stamped application form in hand, Xu Chuan came to the High Performance Computing Center of NTU and found Professor Zhou Qinglin, the director responsible for managing the entire supercomputing center.
"Professor Xu, you are a rare visitor. What kind of wind brings you here?"
Zhou Qinglin, who was at work, was stunned for a moment when he saw Xu Chuan, then quickly stood up and welcomed him with a smile.
"Director Zhou, the main reason I came here this time is to apply for the use of a supercomputer."
After greeting Zhou Qinglin with a smile, Xu Chuan explained his purpose to him.
After listening to his request, Professor Zhou Qinglin looked surprised and surprised.
"You mean that you established a set of phenomenological mathematical control models for the plasma turbulence of a controllable nuclear fusion reactor? Are you sure you're not mistaken?"
To be honest, he, Zhou Qinglin, has been managing the Supercomputing Center of NTU for more than ten years. He has seen a lot of various mathematical models, including mathematical models of fluids, turbulence, plasma flows, etc.
But this was his first time seeing the mathematical model of plasma turbulence in a controllable nuclear polymerization reactor.
Although he is neither a mathematician nor a physicist, he has controlled the supercomputing center for so many years and knows more or less the meaning behind this model.
Xu Chuan shrugged and said, "I don't need to waste everyone's time by joking about this. Your time is precious, and so is mine."
Hearing this, Zhou Qinglin laughed and said, "Professor Xu is joking, your time is much more valuable than mine."
After a pause, he continued: "Did you bring that model over? Let me take a look first."
Taking out the solid-state drive with the model stored in it from his backpack, Xu Chuan handed it over, nodded and said: "Bring it with you, this is it."
Zhou Qinglin took the model, opened it curiously and looked at it. After discovering that there was a hard drive inside, he looked up at Xu Chuan and joked: "It seems that this model is not small, and it needs a hard drive to save it. "
Xu Chuan smiled and said: "The model is not big, but it is probably not that easy to run it. I feel that our supercomputing center may not be able to afford it."
Hearing this, Zhou Qinglin raised his eyebrows and said with a smile: "Then maybe what you need is not a supercomputer, but a quantum computer?"
Although the supercomputing center of NTU is not very powerful and does not rank high in the domestic supercomputing rankings, after completing the upgrade in 2015, it has deployed 910 CPU computing nodes with a total of 21,840 CPU cores. The theoretical calculation peak is 870 trillion times/second.
This level of supercomputing, let alone a mathematical model, is not a problem even if a hundred mathematical models are calculated together on it.
Of course, the model calculations on fluids and turbulence are indeed very complex, but after all, the performance of NTU's supercomputing center is as high as nearly a quadrillion operations per second, so it is impossible that it cannot run.
But he didn't say this. After all, Professor Xu studied mathematical physics and might not know much about computers.
"Come with me, I'll find someone to show you the model first." Professor Zhou Qinglin smiled and walked towards the laboratory of the supercomputing center with the hard drive.
He is only responsible for controlling the supercomputing center and is not a specialized expert.
As for how much computing power the plasma turbulence model created by Professor Xu requires, how many computing cores need to be divided, etc., these still need to be determined through expert judgment.
"Old Liu, come here and help me. Professor Xu has a mathematical model that he wants to test. You can see how to divide the computing power."
In the laboratory, Zhou Qinglin shouted to another middle-aged man who was busy in front of the computer.
After hearing the shouting, the middle-aged man in the laboratory quickly raised his head and looked over.
"Hey, Professor Xu." Old Liu stood up with a smile and walked over, shaking hands with Xu Chuan warmly.
Xu Chuan responded with a smile: "Professor Liu, how sorry I am to trouble you."
"Hey, why bother? Serving Professor Xu, that's what you should do." Old Liu smiled and took the hard drive from Zhou Qinglin. He first connected it to his work computer and scanned it. After confirming that there was no problem, he connected to the supercomputing center.
Click on the hard drive and open the file inside. Lao Liu looked at the data and phenomenological graphics displayed on the computer screen. After thinking for a moment, he said: "Professor Xu, your model is probably not simple."
"If we want to conduct a test, I'm afraid we will have to stop most of the other tasks running in the supercomputing center."
On the side, Zhou Qinglin asked curiously: "Is it so complicated?"
Lao Liu nodded: "This is the most complex turbulence model I have ever seen. It is not only a simulation run, but also contains various modules such as phenomenological energy burst simulation, collision simulation, and magnetic constraints. It needs to be fully calibrated It requires a lot of computing power for testing, and even our supercomputing center may not be able to run it.”
Zhou Qinglin swallowed and couldn't help but look at Xu Chuan.
Good guy, is this a model that cannot even be driven by a supercomputer with a theoretical calculation peak of 870 trillion times per second? Is this something that humans can make? This Professor Xu probably has a quantum computing brain.
After careful inspection and analysis, Old Liu raised his head and looked at Xu Chuan and said, "Professor Xu, if you want to test this model, I'm afraid you need to wait until some of the tasks being executed are finished or can be paused to free up the computing power." alley."
Xu Chuan asked: "How long do I need to wait?"
Old Liu: "Is it urgent? If it is urgent, it will take about a day or two. Director Zhou will contact the school and suspend other tasks first."
"If you are not in a hurry, you will have to wait for about a week. There are a lot of calculations being done now, which will take some time."
Hearing this, Xu Chuan thought for a moment, looked at Zhou Qinglin and asked, "Can you please help me communicate with Director Zhou? Let's test it first. If the model cannot be run here, I can apply for help from the National Supercomputing Center."
Zhou Qinglin said with a smile, "No problem. Anyway, there are no important tasks recently, so it doesn't matter if I pause for a while."
After Zhou Qinglin applied to Nanjing University for communication, the notice and documents for suspending other computing tasks came down quickly.
Although some researchers who were doing calculations complained after receiving the notice, no one said anything after hearing that it was Professor Xu who cut in line.
After all, it's not that you won't continue to calculate, it's just a pause.
In the supercomputing center, after two days of tossing, the various computing tasks that were running were suspended and saved in an orderly manner.
After dealing with these things, Xu Chuan's mathematical model was officially connected to the supercomputing.
In the control room of the supercomputing center, Xu Chuan, wearing a white anti-static suit, watched the staff responsible for processing calculations busy in front of the computer.
Beside them, Zhou Qinglin and Lao Liu stood by and waited curiously.
This was the first time they had seen a mathematical model that simulated the turbulence of plasma in a controlled nuclear fusion reactor. If this model could be successfully calculated, it would definitely be a Nobel Prize-level achievement for the physics community!
"All the current computing tasks of the supercomputing center have been suspended. The supercomputing is currently running well and we can start the experiment at any time."
In the control room, after a busy period, a staff member reported the supercomputing situation to Zhou Qinglin.
Hearing this, Zhou Qinglin glanced at Xu Chuan, who smiled and said, "Professor Zhou, you are in charge. I don't understand this."
"Okay." Zhou Qinglin nodded and turned to order, "Start the test."
After the order was issued, the huge supercomputing center began to operate.
After suspending other computing tasks, the supercomputing center with a theoretical computing peak of 870 trillion times/second took out all its computing power, and the second-copy model was transferred to the pre-assigned permission account in less than a second.
The staff in the control room quickly allocated nodes, cores and other computing resources as required.
Although the scene of more than 900 computing nodes and 20,000 CPU cores working together cannot be seen by the naked eye, in the microscopic world composed of 0 and 1, this is an extremely huge torrent, which has been invincible in the face of any opponent in the past.
But today, they have encountered real opponents.
The plasma turbulence model composed of mathematical formulas and operation rules quickly decomposes into tens of billions or even hundreds of billions of turbulent micro-unit structures after inputting simulation parameters.
These turbulent structures quickly evolve into a doughnut-like operation rule with the joint support of supercomputers and models, running slowly in the supercomputer, simulating the plasma flow in a controlled nuclear fusion reactor.
Looking at the images and data jumping on the screen, Lao Liu, who is responsible for managing the computing power, couldn't help shaking his head.
As he predicted before, the supercomputer of the Nanjing University Supercomputing Center really can't support this model.
The operation was good at the beginning, and the computing power of the supercomputing center was enough to support it, but as the simulation parameters were input as required after the operation started, the output of data became slow.
This means that the supercomputing center needs to spend more time to complete the operation and calculation of the model.
Although the result can be calculated, everyone in the control room knows that this is meaningless for mathematical models that require real-time control.
Turning his head, Lao Liu glanced at Xu Chuan and sighed: "Your model is too complicated, and there are too many calculations for turbulence. In China, perhaps only the top ten national computing centers can test it perfectly."
Xu Chuan stared at the screen and looked at the data on it and replied: "Let's take a look first, just as collecting data."
Lao Liu shook his head and said: "That's the only way."
In the control room, Zhou Qinglin couldn't help smacking his lips as he looked at the calculation data on the screen.
The computing performance of Nanjing University's supercomputing center is not the top three in the country's universities, but at least the top five, but it can't support a mathematical model.
What's wrong with Professor Xu's brain? He actually came up with such a model.
Shaking his head, he looked at Xu Chuan and asked, "If such a complex model is really applied to controlled nuclear fusion, wouldn't it be necessary to build a large supercomputing center for it?"
Xu Chuan shrugged and said, "Why not? As long as it is correct."
Zhou Qinglin was about to say that this was a waste of money. You know, the construction of a supercomputing center starts at more than 10 billion.
But then he remembered that this is controlled nuclear fusion. If it really works, it seems normal to build a supercomputing center to serve it?
PS: (10/10 for the leader Gao Shan Liu Shui. Phew, the boss's extra chapter is finally done. There are still some scattered rewards left. I will write slowly.)
There is another chapter in the evening. Please vote for the monthly ticket.