Rebirth 08: Rise from copycat phones

Chapter 453 Zhiyun Supply Chain: Guaranteeing You Won’t Lose
Some comments from the outside world, whether good or bad, will not affect the progress of the entire X project.

While the outside world was still discussing the X1 test aircraft, Xu Shenxue also learned the latest information about the sixth-generation aircraft project. The partners were carrying out a manned sixth-generation aircraft research and development plan, and finally decided to use the X system as the artificial intelligence system to create a powerful sixth-generation aircraft that combines manned and unmanned aircraft.

The X project team under Zhiyun Group will provide artificial intelligence systems for the two official sixth-generation aircraft models, as well as a series of hardware equipment including onboard computing chips and sensors to maintain the operation of the artificial intelligence systems.

This means that the entire sixth-generation aircraft project has entered a period of comprehensive and rapid development.

Manned command aircraft and unmanned combat aircraft have entered the formal development stage.

In response to this, Xu Shenxue instructed Kong Jiakui, who was in charge of Project X, to do a good job in supporting the project's research and development and not to lag behind.

In fact, Zhiyun Group is not the only company under Xu Shenxue that is involved in the sixth-generation aircraft project. Xiannvshan Holdings is also involved in the project.

However, Xiannvshan Holdings does not provide a complete set of subsystems, but mainly provides some key raw materials and ultra-precision processing equipment... In order to develop semiconductor equipment and consumables, Xiannvshan Holdings has invested huge resources in research and development in the field of materials over the years.

Many new materials and new technologies can not be used in the semiconductor field. Some materials can also be used in other fields.

Even the sixth-generation aircraft project can use many new materials from companies under Xiannvshan Holdings.

The main reason is that materials are too mysterious. People who work with materials are actually not sure whether they can produce anything or what the use of the things they produce will be.

The MMK34 material is a typical example. Yu Taihe originally wanted to develop a new carbon-based material for chip manufacturing, but he never expected that the MMK34 material he developed would have superconducting properties at room temperature and pressure!

So there are so many different kinds of new materials in Xiannvshan Holdings. After picking and choosing, we actually found that quite a few materials can be used in the sixth-generation aircraft project.

Especially last year, Fairy Mountain Holdings developed a carbon/carbon composite material. Like many other materials, this material was originally developed for semiconductor projects. However, after it was produced, it was found that its performance was not very good and it could not be used in advanced semiconductor equipment or consumables, so it was shelved.

There are actually many new materials like this that fail to meet performance requirements and are shelved...

Then Fairy Mountain Holdings and AVIC Engine Corporation of China started to cooperate. The other party asked if there were any new high-temperature materials, especially cutting-edge carbon/carbon composite materials. Fairy Mountain Holdings then came up with more than ten new carbon/carbon composite materials in one go... and almost scared the other party silly!
When did this carbon-carbon composite become so cheap that more than ten kinds can be produced at once?
Well, in fact, they still don’t quite understand the material research and development model of Fairy Mountain Holdings. The material research institutions under Fairy Mountain Holdings have always adopted a parallel research and development model, first finalizing the general direction, and then carrying out a dozen or even dozens of similar projects in one go.

After the approximate routes of these projects are finalized, they are directly handed over to the artificial intelligence scientific research system to run the R&D route to see if the theory can work... If it doesn't work, the project team will modify the design and try again.

If the artificial intelligence scientific research assistance system mode runs smoothly, we will start the next step of specific synthesis and various experiments.

This greatly saves the time and financial costs of material research and development in the early theoretical verification stage, and can even save a lot of experimental time.

Although in most cases, the success rate of doing so is very low, and dozens or even hundreds of attempts may lead to failure...but the problem is that the cost per attempt is also low.

After the R&D engineer has an idea and designs a general direction, the next thing is left to artificial intelligence...Then the artificial intelligence system of the scientific research assistance system has huge computing power, and it runs very fast, so the overall verification cost is actually very low.

With the scientific research support system based on artificial intelligence technology, the materials research of Fairy Mountain Holdings and Zhiyun Group has completely turned into this pure assembly line research and development model...

R&D personnel constantly come up with new ideas and then hand them over to artificial intelligence for verification. If the verification is successful, they will then enter the specific experimental stage.

Therefore, there have been many achievements in various materials in the past two years... Of course, due to the metaphysical properties of materials, most of the so-called new materials are useless. It cannot be said that they are useless, but we don’t know how to use them for the time being.

Not all materials can be the same as MMK34. We were so lucky that after it was produced, the experimenters conducted a special low-temperature conductivity test on it. As a result, they unexpectedly discovered that it has superconducting properties at minus degrees.

If this test had not been conducted at the time, the MMK34 material would most likely have been shelved and left to rot. Perhaps one day in the future someone would remember this humble carbon-based material and, out of boredom, conduct a low-temperature conductivity test, ultimately discovering its superconducting properties.

Materials science, or many other scientific studies, are full of accidents...

Xiannvshan Holdings is a typical example... Many new materials developed are first tested according to the research direction of the R&D team itself. If they don’t work, they will simply test some characteristics that can be used in other semiconductor fields. If they still don’t work, they will ignore them later.

There are all kinds of new synthetic materials coming out every other day. There is no time and resources to waste on some tests that have nothing to do with semiconductors.

Of course, in the future, Xiannvshan Holdings will also look for suitable materials in its own material library based on the requirements of some new projects, and conduct a new round of targeted tests.

The previous engine research and development institution sent a cooperation invitation and asked if there were any new carbon/carbon composite materials. The people at Fairy Mountain Holdings searched through their own material library... and actually found more than ten types.

Of course, it is hard to say whether it meets the requirements. Various detailed tests are required.

After testing, it was found that one of the carbon/carbon composite materials was really good and very suitable for use as the material for some parts of the engine. Not only is it resistant to high temperatures, but its strength under high temperature conditions is also very good, which is better than some of the carbon-carbon composite materials currently used and tested.

They plan to use this new material to manufacture afterburners, tail nozzles and other parts to reduce the weight of the engine.

Because carbon-carbon composite materials are lightweight and high-strength materials, if used in large quantities, they can effectively reduce the weight of the engine and thus improve the thrust-to-weight ratio...and the thrust-to-weight ratio of military engines is the most core performance indicator.

This is also the reason why our partner’s aviation engine R&D manufacturers are looking all over the world for new carbon-carbon composite materials.

Fairy Mountain Holdings' strength in the field of materials does not only lie in carbon-carbon composite materials. It also has outstanding advantages in many other material fields. For example, it is very strong in titanium alloy materials. A considerable part of the body parts of the X1 project team are supplied by Fairy Mountain Holdings, using their self-developed new titanium alloy materials.

Therefore, for the newly developed new engines, many parts will be purchased from Fairy Mountain Holdings in the future, and some parts of the fuselage will also be purchased from Fairy Mountain Holdings.

These are breakthroughs in the field of materials science, bringing about technological advantages.

However, whether it is Xiannvshan Holdings or Zhiyun Group, participating in the entire sixth-generation aircraft project does not actually make much money... Although the entire sixth-generation aircraft project is extremely costly, these funds have to be distributed to numerous research institutions and manufacturers, and Zhiyun Group and Xiannvshan Holdings can only take a part of them.

The most important thing is that the duration of such a huge project is often calculated in ten years, which means that if it is spread out over each year, there is actually not much left.

This year, Zhiyun Group obtained a supporting contract for the X1 project, which will equip the artificial intelligence system for the subsequent four prototypes and the first batch of twelve small-scale test production models of X1 unmanned fighter jets, a total of sixteen aircraft, with a total contract price of billion yuan.

The average unit price of the artificial intelligence system of each X1 aircraft is million yuan, which is equivalent to a little more than million US dollars.

The main reason is that the price of the early prototype and the first batch is actually to subsidize part of Zhiyun Group's huge R&D investment, so the price is so expensive. After subsequent mass production, the price will drop significantly.

It sounds like the price is very high, the total price could reach 11.2 billion, equivalent to a little over 1.6 billion US dollars.

However, the revenue of more than 1.6 billion US dollars in this first batch is just a fraction of a fraction for a giant like Zhiyun Group with annual revenue of more than 400 billion US dollars.

Not to mention the huge cost that Zhiyun Group has invested in Project X. The server array composed of AI series graphics cards used to run artificial intelligence training alone is worth billions of dollars.

The contract amount of more than one billion US dollars is far from enough to cover the initial cost investment of Zhiyun Group... and in order to continue training and upgrading, it will be necessary to continue to invest in computing power costs for training every year, which is very expensive.

If Zhiyun Group wants to avoid losses on Project X, or even make a small profit, it has to rely on subsequent X1 and X3 aircraft, as well as future manned sixth-generation aircraft, to use Project X, and to increase procurement volume in the next decade to at least several hundred aircraft.

Of course, if nothing unexpected happens, there will most likely be hundreds of orders in the next ten years!
Therefore, the possibility of losing money is not great, it’s just that the profit is not much.

Therefore, Zhiyun Group does not actually make much money from Project X. The reason why it is so active in doing it is not simply because of Xu Shenxue's personal opinion, but more because Zhiyun Group, as the country's top high-tech giant, has an unshirkable responsibility in these matters.

Moreover, the various domestic support resources over the past few years have not been obtained in vain.

Since you have taken the resources, you must be aware of the corresponding responsibilities.

Of course, it is actually impossible to really let Zhiyun Group lose money... This company is so big and powerful that it is impossible for them to give much in cash. Even if they know that Zhiyun Group is currently losing money on project X, they will compensate in other ways.

Money is not the only thing that matters!

Other resources can be counted as money!

For example, they will give you a plan to support the development of domestic GPUs, saying that in order to support the development of the domestic GPU field, they hope that newly launched domestic games will support domestic GPU development standards as much as possible... Although they say they hope, in fact, game companies, whether they are overseas or domestic, will basically not give you a game approval number if they do not support domestic GPU development standards.

Don’t ask why. The answer is that it is non-compliant and you have to change it... As for what is non-compliant, just figure it out yourself!
If you want to launch your product in the domestic market without supporting domestic GPU development standards, there is no way...

What is the domestic GPU development standard? Isn't it Zhiyun Group's YGPU standard?

Zhiyun Group has been promoting the YGPU development standard for many years. It was initially promoted in the field of mobile GPUs, and then gradually expanded to PC GPUs and server GPUs. Later, it integrated NVIDIA's CUDA ecosystem.

This development standard occupies the entire market in the server GPU field and is the development standard that must be used for the current development of artificial intelligence... If you do not use this standard, you will not be able to use APO graphics cards, PX computing power platforms and other GPU chip products launched by Zhiyun Group.

The APO graphics card can't be used, so what's the point of playing artificial intelligence...

You can't really use AMD's half-baked AI computing graphics card, because AMD itself doesn't even use that graphics card...

It’s funny to say that AMD also deployed GTAI2 to improve production efficiency, but the graphics card used is the APO graphics card developed and produced by Zhiyun Group... Why?
The reason is simple: GTAI2 is the only open source large-scale generative AI available on the market.

If you want to deploy a local AI model, you can only use GTAI2.

Because GTAI2, it only supports APO series graphics cards!

This means that even if AMD has its own computing power cards, if it wants to deploy artificial intelligence for its own use, it can only deploy GTAI2 and then purchase APO graphics cards.

It's not just AMD. Companies like Intel and Qualcomm have also deployed the GTAI2 large model for their own use, and can only purchase Zhiyun Group's APO graphics cards with reluctance.

This left many people speechless... You know, the above-mentioned semiconductor companies are all direct competitors of Zhiyun Group's semiconductor business.

But as a competitor, they actually have to look to Zhiyun Group to purchase APO graphics cards, which is ridiculous...

This also shows the strength of APO graphics cards, and the strength of APO graphics cards also means the strength of YGPU development standards.

It is natural that China is vigorously promoting the development of YGPU standards.

In addition to supporting domestic GPUs, there will be other compensatory measures, such as approving a "low-altitude aircraft cargo experimental route" for Zhiyun Group.

In a few days, when institutions purchase smart terminals, they will give priority to purchasing Zhiyun phones, Zhiyun computers, tablets and other terminal products and services under Zhiyun Group.

Many times, these contracting projects themselves are not profitable, but they can obtain sufficient support and profits from other aspects.

Project X undertaken by Zhiyun Group is a typical example... After Project X was successfully advanced and the formal contract was obtained, Xu Shenxue also paid less attention to it. The next Project X will continue to be advanced step by step.

On the one hand, we will cooperate with our partners to first manufacture several prototypes in the X1 project, and make targeted improvements based on the data collected from the prototype test flights, in order to prepare for the subsequent first batch of twelve small-scale trial production models.

At the same time, we must cooperate with research and development institutions to modify the existing two-seat model of the fifth-generation aircraft to carry a simplified version of the X system to control and command the X1 UAV.

On the other hand, we will continue to adapt the X3 project and continue to improve and adapt their prototypes to produce an X3 fighter version of the X project artificial intelligence system.

Finally, we will cooperate with them to develop manned sixth-generation fighter jets.

The entire project will progress steadily. If everything goes well, perhaps in two years we will be able to see the first X1 project production model, that is, the first of the first batch of twelve aircraft to roll off the production line and make its maiden flight. Of course, if things do not go well, it may take three or even four years to enter small-scale production.

Whether this can proceed smoothly is not just a problem of the X project itself, but more about the technical difficulties in the body, engine and other areas.

The X1 project, a sub-project in the sixth-generation aircraft project, has a very high technical content, especially the technical content of the completed body.

The two prototypes manufactured previously actually failed to meet the design goals and were just temporary makeshift tests.

For example, the most core engine is not very good. The planned engine is a new generation medium-sized engine with a thrust-to-weight ratio of more than ten, but the current engine is still an old generation of imported engines, and its performance does not meet the requirements at all.

Although the planned engine has a very fast development progress due to the technical bonus of many new materials and new processes of Zhiyun Group and Xiannvshan Holdings, it still takes time to make it usable. Even though a prototype is already available, it will still take several years to reach mass production.

It would be ideal if we could complete it within two years, but it might take three or four years.

This is even more true for the X3 project, which also requires a new generation of engines, and this engine is still in the research and development stage... This engine is actually also the pre-ordered engine for the sixth-generation manned aircraft. Its technical content is very high, and it is not easy to produce it.

Since the X project was officially launched in 13, it took four years to build the X1 prototype for technical verification and test flight.

It is estimated that it will take another two to three years to enter small-scale production, and the entire R&D phase is expected to take seven to eight years or even ten years.

And this...actually does not include the early theoretical and technical research and development stage of the cooperating airline. If the preliminary research stage is truly included, the time span can reach more than ten years.

But even so, it is actually very fast.

The United States had a corresponding plan for the sixth-generation aircraft in 09, and a formal project team was established a few years ago, but now even the design direction and PPT pictures have not been worked out.

------

After Xu Shenxue temporarily put aside his attention on the sixth-generation aircraft project, he turned his focus back to the field of intelligent robots.

Mainly after the release of the new generation Yun Robot B3 at the end of December, Zhiyun Group once again ignited the wave of intelligent robots.

Compared with the previous generation Yun Robot B3, Yun Robot B1 is more advanced, has a wider range of applications, and its overall performance has more than doubled. This has greatly boosted the attention of the entire intelligent robot market, and even significantly increased the sales of the two entry-level models Yun Robot C6/C7.

At the same time, the old generation of B1 models has experienced a year of price cuts and now only costs about to yuan, and its sales have also increased slightly.

The B1 and B3, as well as the C6 and C7 in the Yun Robot, these four new and old models with different prices together bear the heavy responsibility of sales for Zhiyun Group's intelligent robot business.

The rapidly increasing orders have also posed a huge challenge to Zhiyun Group’s intelligent robot production capacity!
Weiku Industrial and BYD's OEM factories that produce intelligent robots are now operating 7/24. Workers work in three shifts, maximizing the capacity of existing production lines.

At the same time, the construction of new factories is also being accelerated, striving to complete it and put it into production as soon as possible.

Zhiyun Group wants to achieve a production capacity of 1.5 million intelligent robots next year, which is not an easy task.

The intelligent robots in Zhiyun Group are much more difficult to manufacture than the smart electric cars in Hailan Automobile.

Not only is the assembly of the finished product complex and technically difficult, the most critical issue is that the technical difficulty of the various components is extremely high, and the supply chain manufacturers' production of spare parts is also limited.

Among them are the typical ZY34 and 35 chips, as well as the EYQ4 chip, both of which are chips made using ten-nanometer technology.

And they are all produced using advanced 3D packaging technology.

This means that to produce these three chips, it is necessary to compete with the current S-series chips for the 6000-nanometer process production capacity, and to compete with the AI4600 graphics cards and APO3 graphics cards for the D packaging process production capacity.

Both types of production capacity are extremely scarce.

In addition, there are dense sensor arrays in the robot's hands and feet. This thing also has a very high technical content. The production and manufacturing of the world's first-class semiconductor sensors requires very excellent process technology. It needs to be produced by Zhiyun Microelectronics' professional advanced semiconductor sensor factory. However, this top-level semiconductor sensor manufacturing process has very limited production capacity.

In addition to these core items, many other spare parts are also top-level technological products, and they are all customized spare parts. The supply chain manufacturers do not even have ready-made products, and new production lines are needed to produce them.

To be honest, if it weren't for Zhiyun Group's orders, these supply chain manufacturers would not dare to invest huge sums of money to build these top-level production lines.

The various equipment and even technicians in this top-level production line are very expensive!
Moreover, the products produced are only needed by Zhiyun Group’s intelligent robot business. If the robots don’t sell well enough, they will lose all their money!
In other words, Zhiyun Group has a great reputation, and the supply chain manufacturers have a blind trust in Zhiyun Group, which is why they invested huge amounts of money in the early stages to carry out supporting research and development and production line construction specifically for intelligent robots.

However, even so, most of the current supply chain manufacturers are still in a loss-making state in the intelligent robot parts business... In order to promote intelligent robots, Zhiyun Group continues to reduce the manufacturing cost and selling price of robots.

Part of the pressure to reduce costs is borne by Zhiyun Group, especially some core products such as computing chips.

However, a large number of supply chain companies are also under tremendous pressure to reduce costs... Gu Zhiming, senior vice president of Zhiyun Group, has been wielding a whip every few days, constantly beating a group of supply chain companies, forcing them to find ways to significantly reduce costs through technological advances.

As for reducing costs by exploiting workers, this doesn’t make much sense... because labor costs are only a small part of the cost of these parts, and the bulk of the cost still lies in technology costs.

Zhiyun Group requires supply chain companies to reduce costs by a certain percentage every once in a while.
The problem is that the quality requirements for spare parts are still very strict and will not be compromised at all.

In order to provide support for intelligent robots, supply chain manufacturers have to pay very high initial research and development and current production costs.

If they only rely on the current sales of hundreds of thousands of spare parts, they will not be able to recover such a huge investment cost... They have to rely on more subsequent orders.

Many supply chain manufacturers quote spare parts based on the assumption that Zhiyun Robotics will be able to reach an annual production capacity of two or even three million units in the future... This is the only way to spread the current average cost. If not, the price of each spare part will be sky-high!

Just like the original prototype, the initial estimated hardware cost exceeded 800,000 yuan, plus the costs in other manufacturing processes, it would definitely be over one million yuan... There is no way to commercialize it, it is too expensive, and sales will be very limited. It is estimated that at most tens of thousands of units will be sold a year.

It’s not about whether the robots are good or not, but that they are too expensive and few people can afford them!
Therefore, in order to promote intelligent robots, Zhiyun Group negotiated with a large number of suppliers, persuading them to lower the prices of spare parts first, so that I could reduce the cost of robots first, and then launch them into the market and detonate the market.

As for making money, wait until the market explodes in the future!
Only Zhiyun Group can say this. If any other company said this to these supply chain manufacturers, they would have been kicked out long ago.

But Zhiyun Group... is an exception!

Therefore, these supply chain manufacturers made up their minds and agreed to Zhiyun Group's conditions, suffering losses for a period of time, and then making profits after developing the market.

So sometimes, it is not easy to be a supply chain company of Zhiyun Group. Not only do you need good technology, but you also need strong financial strength and the ability to endure loneliness. Only in this way can you follow Zhiyun Group to open up new markets and make a lot of money in the future.

The supply chain of Zhiyun Group is not easy, which is also reflected in the purchasing price and technology research and development.

There is a characteristic of Zhiyun Group's procurement, that is, the unit price given for the initial order is often very good, but after the first batch of orders, the subsequent orders will be reduced after a period of time.

This will force supply chain companies to actively reduce costs while developing new generations of technologies.

And these are the costs!
With such a huge investment, a group of supply chain manufacturers in the intelligent robot business are now under great pressure... With so much money invested, if Zhiyun Group's robot business fails, many supply chain manufacturers will go bankrupt!
So much so that some people believe that Zhiyun Group is now engaging hundreds of supply chain manufacturers and hundreds or even thousands of indirect supplier companies behind them in a huge gamble: betting on the future of the intelligent robot market!

But can Zhiyun Group win?

No one knows for now!
No matter how good the smart robots of Zhiyun Group are, the price is there, and it is impossible for them to be cheaper.

Furthermore, smart robots have another characteristic, which is that their usage scenarios are still relatively limited. They are still mainly used as home smart robots, and are still aimed at high-net-worth groups, and these people are few in number!

Therefore, many people have doubts about how much sales of intelligent robots will be in the future!
The Hailan SLC is good. It has basically all the configurations that can be installed. Moreover, the whole series is equipped with L4 automatic driving, large-capacity semi-solid battery plus extended-range engine technology, and various other top-level spare parts. It also provides a bulletproof version...

But as for sales volume, the orders for it are only about 10,000 vehicles a year!

It’s not that it’s not good enough, but that it’s too expensive and too few people can afford it!
The situation is similar in the intelligent robot market. High costs and high prices mean that there will be no sales... and no sales will in turn push up the average cost of each robot, creating a vicious cycle.

Therefore, Zhiyun Group adopted another method. They first reduced prices at a loss, then lowered the selling price to increase sales. Once sales increased, costs in all aspects would naturally be greatly reduced.

When the data is summarized and calculated, all the previous losses can be earned back, and there will be a good profit.

This is the first to break the vicious cycle!
The only one that can get a large number of supply agencies to do this is Zhiyun Group!
Who made it so awesome!

Besides, Zhiyun Group doesn’t lack their money, and the external procurement orders are huge!

Even if Zhiyun's robot spare parts business suffers losses, Zhiyun Group will be able to make up for this loss with spare parts purchase orders from other smart terminal businesses!
To put it bluntly, Zhiyun Group has sufficient capabilities to provide support for supply chain manufacturers!

Gu Zhiming was almost telling the supply chain manufacturers: "Just go for it, spend the money and recruit the talent!"

"For this robot project, if it succeeds, it will be a big profit, but if it fails, it will be a small profit!"

"Even if I fail completely in the end, I can still make it up to you in other ways."

"We at Zhiyun don't have much else, but we have a lot of orders!"

"Anyway, feel free to do it. I, Zhiyun, will ensure that you will not lose anything!"

After Gu Zhiming's deception and threats and inducements, a large number of supply chain manufacturers' bosses sold all their belongings and followed Zhiyun Group to develop intelligent robots...

(End of this chapter)