Black technology starts from the steel suit

Chapter 187 187 Applications of Massively Parallel Computing
Large-scale parallel computing can almost be seen as a "Cambrian" explosion of a new computer architecture in the next decade. At that time, computer architects in academia and industry will usher in an exciting era.

Parallel computing is the basis for all high-performance computing and the exponential explosion of AI computing power. Cutting-edge technology applications such as AI, ML, quantum computing, and biological computing all rely on parallel computing for support.

Parallel computing is now in a whole new era. Computer architecture masters Hennessey and Patterson, who won the Turing Award in 2017, pointed out that whether it is in instruction-level processing technology or multi-core processing technology, the inherent low efficiency of general-purpose processors and the end of the law of scaling ratio and Moore's Law make the processor architecture Designers and designers are unlikely to maintain significant performance improvements on general-purpose processors. "The improvement and innovation of computer architecture must be carried out simultaneously with the improvement and innovation of parallel algorithms and parallel software. The higher the level of improvement, the greater the improvement in performance and efficiency. Whether it is breaking through Moore's Law or von Neumann's bottleneck, it must across levels."

In the process of epidemic prevention and control, the team led by Academician Li made significant contributions to the anti-epidemic and disaster relief in terms of large-scale parallel computing.

****Suddenly, the application of advanced computing technologies such as high-performance computing and artificial intelligence has played an important role in virus research and epidemic prevention and control.With the boost of the recent "new infrastructure" tide, parallel computing will truly usher in a "golden age" in the next few decades.

Because the development of new vaccines and antiviral drugs involves complex mathematical models, numerical methods must be used to solve them through parallel computing. Parallel processing technology will play an irreplaceable role in the fight against viruses.

The same is true for molecular dynamics simulations. The parallel computing model is almost tailor-made for this kind of simulation operations with huge requirements for computing power and computing performance.

Unfortunately, however, too little research has been done in this area.There are many reasons, two of the most important are:
One is that the effect of molecular dynamics simulations using supercomputers is not particularly good at present, and the experimental results are not accurate. With this effort, it is better to do experimental research to get more accurate and more valuable data;
The second is that the simulation time is too short. For example, you can simulate a second of his collision with a supercomputer for a month. It’s cheap, so let’s take an example this time. Although both of them are not short of money, they burned more than 200 million US dollars in just three short experiments, and nothing has been done yet.

If you want to do an experiment, even if it is burned, there will still be a little ashes left, but after a failure occurs during the simulation operation, the remaining data in the computer is basically of little value.

Therefore, although Academician Li is an expert in this field, in reality there are not many opportunities for him to show his talents.The support an academician can get is still too little, at least compared to David Xiao and people like Wang Feng who are very, very rich and well-connected, the resource support they can get is not enough.

But this is obviously a thing of the past. With Wang Feng's financial support, coupled with the existing technical information of David Xiao's technical team, and the full support of a professional semiconductor design team like HiSilicon, Lao Li will not have any problems. Such a rich battle has been fought!

Standing in the laboratory, Academician Li can be said to be very emotional!
Although this time he accepted his invitation, on the one hand, it was to give his mentor some face, after all, everyone bowed their heads and did not look up.But this is not the most important thing, the most important thing is the opportunity provided by the opponent to flex its muscles, which is really too precious.

Although our annual investment in scientific research has begun to increase substantially, there are also many people who want to share the money. On average, there may not be much money to give to one person.Moreover, as the level of education continues to improve, more and more people enter this industry, and it is naturally more and more difficult to distribute money.

Although Academician Li has a high enough status and seniority, he still has to take care of everyone's feelings, so if he wants to do something big and grand, he basically has no chance.So after hearing Academician Zhu's invitation, he came over with just a little reserve: Although he was quite famous in this incident, there are many others with similar levels to him. Wouldn't it be a big loss for one chance?

fine…

Fortunately, he agreed early. To be honest, he was stunned at the time. He knew that one of these two guys is a fund and the other is a new energy company. right?

Good guy, let's not talk about hardware.The team composed of elites drawn from the top five semiconductor design teams in the country alone was something he could not have imagined before.One is that everyone belongs to different departments, and he doesn't have such a big face yet.The other is that they want to use such a large team, which means that the investment this time has gone to the sea.Among other things, according to his understanding, the chip this time is estimated to be made with a seven-nanometer process.You must know that the design cost is not mentioned first, the cost of a tape-out alone is as high as several hundred million, which is not something ordinary people can afford.

With such hardware conditions and technical support, Lao Li felt that he could talk about teenage madness, and he could not miss this opportunity.If you don't rely on such an opportunity to make yourself famous, how can you have a second time?
To be honest, he has not had the experience of laying a floor in the laboratory for a long time. After all, one is that he is old and his physical condition does not allow him. I am a young man who is admired by my mentor, so there is no need for it.

But now, for his original dream, he spent an astonishing 14 hours on average in the laboratory in the first week. It can be said that he spent all his time except for eating, sleeping, and going to the hospital for examinations. lab.

This scene surprised the students who are familiar with Lao Li's habits and other researchers. My boss, what kind of stimulation is this? Is it because Wang Feng, a young man, doesn't talk about martial arts, relying on his own identity and status? , bullying Lao Li, an old man?
But it shouldn't be, Lao Li didn't look unhappy at all, on the contrary, he enjoyed it.How do I feel about it? It’s like when they were young, they followed the middle-aged version of Lao Li. At that time, Lao Li’s body was still very good, and he was not as tall as he was. No matter what he did, he had great energy and enthusiasm. .

But now...they remembered what the teacher said: You must supervise Lao Li and let him go to the hospital on time.

(End of this chapter)