Chapter 534: The semiconductor industry chain must be started from the beginning

Today's dinner party is a good occasion for socializing, but there are many people and there is not much time to talk about too many things in detail.

Moreover, even if it was just a formality, Zhu Lan had to ask Lin Guofan for his opinion.

Not to mention that Chang Haonan is just cooperating with others, the relationship between the two parties is equal.

Even if the superior unit wants to transfer work, except for a few special circumstances, it must first seek the consent of the family members.

Therefore, after initially confirming the intention to cooperate, the more specific content must be discussed later.

After Zhu Lan and Chang Haonan said goodbye and left, Li Jie, who was sitting next to him, looked at the former's back and asked in surprise:

"Comrade Xiao Chang still has such familiar colleagues in the northwest?"

Those invited to today's awards ceremony will wear badges with their names and workplaces written on them.

After the ceremony, most people didn't have time to pick it up and were taken directly to the banquet hall on the second floor.

That's why Li Jie could directly tell that Zhu Lan was from Jincheng University in the northwest.

"Ah, we met when we were doing aviation meteorological research in Nanzheng. Professor Zhu and his husband Professor Lin Guofan made a lot of efforts at that time."

Chang Haonan certainly has nothing to hide in these contents.

Li Jie paused for a moment, and then said as if talking to himself:

"If we really want to build the Qinghai-Tibet Railway, Jincheng University should really contribute more..."

Of course Chang Haonan knew that this railway must be built, but at this time he still had to pretend:

"Does the country really have plans to build railways in snowy areas?"

Li Jie nodded:

"It has been around for a long time. The Xining to Golmud section started construction in the 1950s, but it took almost 30 years to complete and open to traffic. At that time, it was felt that further repairs were too technically difficult and not feasible, so it was temporarily shelved. "

After briefly explaining the situation to Chang Haonan, he changed the subject and asked the latter:

"What you just said is to use numerical calculation methods to analyze the situation after dynamic loads are applied to the roadbed in the frozen soil area. Are you really sure it can be done?"

"At least from my professional point of view, there is no problem. The extension toolkit for the infinite element method will not take long to be developed."

Chang Haonan gave an affirmative answer almost without thinking:

"Of course, any numerical calculation that is applied to specific engineering problems requires corresponding data support, and it is impossible to calculate in a vacuum."

"So there are related characteristics of frozen soil, such as convection, mass migration, water evaporation heat consumption, heat conduction between soil skeleton and other media, ice-water phase change, and of course other characteristic data that geological experts know but I don't know. It is still necessary to conduct research and analysis on the actual situation along the route..."

As a researcher in the field of numerical computing, Li Jie certainly knew that what Chang Haonan said later was a supplementary explanation out of rigor.

So he directly condensed the entire answer into one word——

can.

"If this is the case, then our Shuguang Group should prepare in advance..."

In addition to being the new director of the Computing Institute, Li Jie is also the chairman and general manager of Shuguang Group.

Due to the butterfly effect set off by Chang Haonan, the situation of the Institute of Computing and Shuguang Group was much better than in the previous life, at least they were not reversely annexed by the Dream Group, which now operates independently.

However, as an asset-heavy R&D company, the latter needs to work harder to maintain a balance of payments.

Chang Haonan is certainly aware of this situation.

So when Li Jie said this, he also became interested:

"Speaking of this, Academician Li, I wonder how the research and development progress of Dawn 2000 is going?"

"The current computing speed of Dawning 1000 and Dawning 1000A can no longer keep up with our needs."

Facing Chang Haonan's strong reminder, Li Jie briefly calculated the current timetable, and then replied:

"If nothing else happens, Sugon 2000-1 will be put into use in the second half of this year. Based on some experience of previous Sugon 1000A feedback, the calculation speed of the new model can be increased by about an order of magnitude."

“What I’m thinking about now is whether we should design a suitable supercomputer for very large projects like the Qinghai-Tibet Railway in advance, so that when the project is actually approved, even if the performance of our product is slightly inferior, we don’t have to wait for it to be in stock. After all, we can still get some usage share..."

Computer hardware is actually similar to software. We are not afraid of falling behind temporarily, but we are afraid that there will be no market.

Without a market, there would be no feedback, and it would not be able to bring about an entire industry chain.

The wave of supercomputing craze that Chang Haonan set off last year can be regarded as solving the urgent need of Sugon Group.

Otherwise, the capital chain may become unstoppable.

But just relying on a few universities, and these are all orders from second- and third-rate universities with poor conditions, is definitely not an option in the long run.

Dawn series supercomputers still need to prove themselves on a larger stage.

But if we really want to surpass foreign countries, especially our counterparts in the United States and Japan, in hard indicators, it is obviously not something that can be achieved within a year or two.

That's why Li Jie thought of taking some sideways paths.

I have to say that this idea is indeed something Chang Haonan has never thought of before.

But after hearing Li Jie's idea, Chang Haonan's mind suddenly came to an idea of ​​​​more freedom:

"Academician Li, I think we should do it more simply."

"Can Sugon Group cooperate with our Torch Group to optimize a supercomputer specifically for the TORCH Multiphysics numerical calculation software? For example... it can be called Sugon 2000T and specially provided for use in very large projects like the Qinghai-Tibet Railway. Anyway, they It’s a dedicated aircraft, so you don’t have to worry about compatibility…”

He didn't know much about the computer hardware world.

But judging from past experience, it should be feasible for computers to optimize specific types of applications from the hardware, or at least the underlying driver software level.

For example, it is better to use N card for mining than A card...

Now, it was Li Jie's turn to shine:

"You mean...bundled sales..."

"No...don't say it so harshly..."

Chang Haonan quickly waved his hand to deny:

"Can we call it a joint model..."

"Forced buying and selling is definitely not good, but we can add some exclusive functions of the Sugon 2000T co-branded model to the professional version of the software for the public. After unlocking it, it is equivalent to purchasing half of the professional upgrade model, and then we recommend that customers choose Sugon. The group’s products…”

"Of course, part of the profits from this joint model must be shared with our Torch Group..."

Chang Haonan's gameplay is somewhat similar to a certain platform exclusive in the gaming industry, or at least a first-release exclusive, which is a very common marketing strategy in later generations.

However, in China in the late 90s, it was obviously still relatively new, which at least made Li Jie feel a little brain-burning for a while.

After a while, he came back to his senses:

"If we want to do this, it will not only be a problem of the system and architecture, but also the core hardware, such as the chip level, must be modified accordingly..."

"Of course we have to change it."

For Chang Haonan, who returned from time travel nearly twenty years later, "making independent chips" is almost an instinctive idea.

So, he continued to explain:

"If our supercomputer, and even the entire semiconductor industry, wants to develop, we cannot always rely on finished hardware provided by foreign countries. One day we must rely on ourselves."

"Since such a day will come, it's better to prepare early."

"It's not like we need to get it right in one step. We can start with the design first. At worst, we'll have to hire someone else for production. Supercomputing chips are not consumer-grade products and are relatively less sensitive."

"In short, we not only want to revitalize Shuguang Group, but also pull up the upstream industrial chain bit by bit, at least not letting people fall further and further behind."

"Right now, the international environment is relatively not that tense, and there is still the possibility of obtaining some technology and collaboration from abroad. If we don't take advantage of this opportunity to start pursuing it quickly, if the situation changes in the future and we want to break through, it will be very difficult. Too much..."

Clear and logical.

Li Jie took a deep breath, stared at Chang Haonan for a while, and then nodded:

"Okay, I'll make arrangements when I get back."