Chapter 301: Nantianmen Series Chips(1/2)
Chapter 301 Nantianmen Series Chips
Having said this, Feng Rong suddenly became very interested.
The two of them looked at each other with eager expressions in their eyes.
"Lao Zhang, Lao Feng, let's go and see for ourselves!"
After saying this, Hu Lai stood up and took Feng Rong and others to the Phoenix Satellite Department.
Soon, the four of them drove to the Phoenix Super Factory.
The newly established satellite department is located in the Phoenix Super Industrial Park. Unlike other large departments, the satellite department only has a R&D laboratory and technical team.
The production and manufacturing of the satellites are all produced by the "Happy Planet" company, and the equipment on the satellites are basically purchased from other domestic parts manufacturers.
To put it simply, the Phoenix satellite department is an "asset-light" department, which is also in line with the current commercial satellite market, which is not large and the principle of "the stall does not need to be too big."
Director Zhang led the way. The four of them had just walked into the office building of the satellite department when they turned the corner and bumped into Zhang Xiaoou, the "Happy Planet" whom they had just met two hours ago!
When he was stunned, Director Zhang was the first to react.
"Mr. Xiaoou, are you there too?!"
Zhang Xiaoou was also stunned for a moment, then greeted Hu Lai and others warmly with a bright smile and explained:
"Mr. Hu, Mr. Zhang, Mr. Zhou Damin called me just now and asked me to come over and participate in the project discussion of our two subsequent 'viewing series' satellites. The project team should finalize the plan today."
"Mr. Hu, you came just in time. You are in the conference room. Let's go there together."
The person Zhang Xiaoou mentioned was Zhou Damin, who is now the main person in charge of the satellite department. He was a scientific research and academic expert specializing in satellites recruited by the Phoenix Research Institute when it established the satellite research and development project.
Now that the Phoenix Satellite Department has been established, Zhou Damin naturally becomes the head of the satellite department.
After a few simple exchanges, Zhang Xiaoou walked in front and led the crowd. At this moment, Hu Lai suddenly noticed that Wang Lulu in the crowd took two hesitant steps back.
Hu Lai was stunned for a moment, then immediately said with a smile:
"It doesn't matter, Sister Lulu, we are not outsiders."
"If you are interested, you can listen together. If you are not interested, I can arrange for someone to accompany you around our industrial park."
Wang Lulu felt warm in her heart. She had indeed considered that such a meeting was a business secret of the Phoenix Satellite Department and it was not appropriate for her to participate.
Unexpectedly, Hu Lai's observation was quite meticulous. He didn't show any pretense at the moment and said calmly:
"Since Mr. Hu doesn't care, then I've learned a lot."
"Okay, let's go listen together."
Hu Lai nodded, and then the group walked towards the conference room.
As soon as Hu Lai and everyone arrived at the door of the conference room, they heard bursts of heated discussions coming from inside.
"General Damin, we have completed the 'AI intelligent recognition' function that you want to implement on your satellite according to your request, but I still recommend that you think about it carefully."
"According to your requirements, it is best to have an AI chip equipped with 100-400 TOPS of computing power to play the best role of the 'onboard AI intelligent processing system'. Your current solution is only equipped with an AI chip with a computing power of 40TOPS. I think
Too little, too conservative!”
"and."
"Based on my understanding of commercial remote sensing satellites these days, as sensors become more and more advanced in the future, the amount of data that needs to be processed on satellites will also increase. It is estimated that chips with 40TOPS computing power will not be able to keep up with the future in the near future.
The development trend of commercial satellites may be eliminated in just a few years!"
The voice in his ears was familiar. Hu Lai was slightly startled and walked into the conference room. He immediately saw Zhou Xinxin who was talking.
Zhou Xinxin is an expert in artificial intelligence. He developed the smart cockpit and artificial intelligence robot of Phoenix Automobile's "Qingluan". Unexpectedly, Zhou Xinxin also came to the satellite department meeting today.
but.
When Hu took a look at the people in the conference room, he discovered that not only Zhou Xinxin, who was in charge of artificial intelligence, but also Lao Wang, the technical director of the Phoenix chip department, was also in the conference room.
Unexpectedly, today's satellite project meeting brought together the artificial intelligence department, chip department and "Happy Planet" Zhang Xiaoou.
Everyone was quite surprised by the sudden visit of Hu Lai and others, and greeted Hu Lai and others one by one. After being polite, Hu Lai curiously returned to the topic Zhou Xinxin had just talked about.
"Xinxin, what do you mean by what you just said about Damin being too conservative and the 40TOPS computing power chip not keeping up with the development trend?"
After Zhou Xinxin explained it, Hu Lai quickly understood the cause and effect.
As Director Zhang said, the project team is preparing to use advanced high-performance 7-nanometer computing power chips for the first time on "Guanjing 2" and "3".
After being equipped with a high-computing power chip, the project team is preparing to use the high computing power of the AI chip to build an "on-board AI intelligent processing platform" on the satellite that can perform intelligent processing, turning the "viewing satellite" into a smart satellite
.
The 40TOPS computing power that Zhou Xinxin just said is too little actually means that the AI chip used by the project team on the satellite has too little computing power, so I specially reminded you.
After figuring it out, Hu Lai was not in a hurry to speak. He had the impression that Phoenix's first chip, the "Tiger Leap 140" mounted on the car, also had a full 120TOPS.
Why, the 7nm satellite chip only has 40TOPS?
Before he could speak, Zhou Xinxin looked at Zhou Damin and continued:
"So, Mr. Damin, I suggest that you use larger computing power chips for satellite applications. It is best to adopt the method of 'hardware pre-embedding' and 'software upgrade'!"
"This method is widely used in the new energy vehicle industry. Whether it is Xiaopeng, Weilai or our Phoenix 'Qingluan', they are all using it. Just like the Tigeryue 140 chip equipped with our 'Qingluan', whether it is used or not now
The computing power is 120TOPS, but if the hardware performance is improved in advance, no matter how the software is upgraded in the next five or six years, the hardware will be sufficient!"
"This is especially true for your satellite. After all, it is a disposable item once it is launched into the sky. It is impossible to repair and replace the hardware."
Zhou Xinxin has indeed been working on intelligent AI robots at Phoenix Automobile for a long time, and the ideas in his mind directly transplanted the way new energy vehicles deal with future development.
Hu Lai carefully considered Zhou Xinxin's words. Commercial satellites are currently in a period of rapid development, which is indeed similar to the development of new energy vehicles.
This kind of "hardware pre-embedding" method can indeed extend the service life and expand the functionality of the satellite as much as possible.
but.
After the words fell, Feng Rong on the side suddenly said:
"I'm afraid it's not that Zhou Damin doesn't want to use high-computing AI chips, it's probably because it's difficult for existing chip packaging technology to balance stability and performance in a space environment?!"
"Haha, that's Lao Wang's problem."
"General Damin is so easy to talk to, so you should only be responsible for making requests and leave the rest of the problems to Lao Wang! He is now in charge of the technology research and development of the entire chip department. If you don't put some pressure on him, how do you know he can't do it?
!”
Zhou Xinxin said with a smile, obviously he also knew that it was difficult to make aerospace-grade high computing power chips.
However, after he finished speaking, he still glanced at Feng Rong, who was not very familiar with him, in surprise.
Just when everyone was making fun of him, Hu Lai also reacted.
Simply put, electronic equipment systems are important supports for a series of spacecraft such as satellites, Mars probes and even space stations, and aerospace-grade chips are the core of aerospace electronic equipment!
These aerospace-grade chips have been programmed with various predetermined programs by ground researchers before departure, and will be executed according to the predetermined programs when triggered.
It can be said that aerospace chips almost determine the success or failure of aerospace missions.
But the difficulty is that there are not only extreme environments of high temperature and low temperature in space, but also a lot of high-energy particles and cosmic rays. These particles and rays will penetrate the shielding layer of the spacecraft and interact with component materials to produce radiation, which in turn will cause
Component function degradation!
Therefore, in order to ensure the reliability of the chip, the spacecraft used in space missions in recent years have basically used chips using low-end process technology.
The reason why a large number of chips of 90 nanometers or even above are used is because the larger the nanometer number, the more reliable the chip is and the smaller the chance of problems.
but.
As we all know, the larger the nanometer of the chip, the worse its performance. This has also led to the fact that the chips on satellites are basically at the level of consumer-grade chips 20 years ago.
Thinking of this, Hu Lai also understood that the difficulty of developing aerospace-grade chips and consumer-grade chips was very different.
If the Phoenix Satellite Department wants to develop smart AI satellites, it must solve the problem of aerospace-grade high computing power chips.
At this moment, the voice of chip director Lao Wang came to his ears. He looked up and saw Lao Wang looking at Zhou Xinxin with an angry look.
"Well, well, Zhou Xinxin, you don't think it's a big deal."
"Do you know how much power consumption Damin always gave us to limit the chip?"
"A satellite chip with 12 watts of power! What else do you want with 12 watts of power? Do you know that even Qualcomm's Snapdragon 8 GEN 1 chip now has 11.5 watts!"
"Not to mention that this is not a consumer-grade chip, so you don't have to worry about various scenarios, this is an aerospace-grade chip!"
“How much effort have we put into isolating particle encapsulation before we can barely achieve some results, we can barely achieve the level of 40TOPS at 7 nanometers!”
"Besides, I'm not 100% sure it's stable yet. I still need President Xiaoou and the others to do some special radiation absorption on the satellite's protective layer!"
Zhou Xinxin smiled nonchalantly and said nothing.
After a while, Zhou Damin, the head of the satellite department who had been silent, coughed slightly and said slightly embarrassed.
"Our satellite project department has nothing to do. Who doesn't want to leave more power for the chip?"
"But we are small satellites after all, and our energy sources almost all come from solar energy. These energy are not only used for electronic systems, but also provide energy output for various photographic instruments carried by the satellite."
"Besides, there is a vacuum in space and there is no wind. Once the power is increased, we will have to redo the entire satellite's cooling system. Then everything will have to be changed!"
"So... 12 watts is already the limit we can spare. For the last 'Guanjing No. 1', we only used 4 watts of power consumption chips!"
To be continued...