These days, Huawei ’s self -developed Kirin chip is still in its infancy, accounting for only 30%-40%of the total number of Huawei chips.
Wang Yifeng turned: "In the past, Huawei used Qualcomm chips, MediaTek chip. You also talk about selling our Kunpeng chip to them, try to achieve in -depth cooperation.
Point discount. ""
"Okay, this is a win -win!" Zhang Chao agreed.
After all, Huawei Buy Lianfa and Qualcomm's chips are bought, and the chip of Xingyi Technology is also bought.
"Yes, a win -win situation, we win twice! Over the founding Kirin chip once, sell Kunpeng chip to Huawei again." Wang Yi agreed.
Zhang Chao's mouth twitched: "..."
Win -win ... Win twice ... God logic!
But think about it, nothing wrong.
Subsequently, Zhang Chalei became popular, and he moved to the city that day.
As for the research and development of Kunpeng 906 SOC, all verification work has been completed.
At present, the R & D factory of the wafer factory is preparing for flowing slices, so it is waiting for the flow at the end of the month!
These preparations are mainly chip production. Zhang Chao is not a problem. When the slices are coming, it is present.
At the same time, Huawei also held a high -level meeting to discuss related matters.
Lao Yu took the lead: "Regarding the chip foundry problem of the Kirin 910, there are currently two options. One is TSMC, which was originally planned. Our K3V2 is their foundry."
"But recently Xingyi semiconductor has extended an olive branch, thinking about giving us foundry. And the price is more favorable."
"So how do we choose?"
Everyone is in contemplation and different thoughts.
"What do you think, even though you say it." Yu Dong said with a smile.
Everyone responded and spoke:
"I suggest still use TSMC for foundry. TSMC as a semiconductor giant, with richer experience and more mature craftsmanship."
"I oppose that although Xingyi Semiconductor is an emerging semiconductor company, it has improved and has a rapid future. Their 28 nanometer process is excellent, not worse than TSMC."
"The 28 -nanometer of Xingyi Semiconductor is very good, but the question is the next?"
"TSMC and Samsung are developing 20 nanometer technology, and the 20 -nanometer process can be promoted next year. We use TSMC OEM, and we can also use 20 nanometers in the future, but what about Xingyi Technology?
The craftsmanship is still far away? "
"This is a big problem. Once we abandon TSMC and choose Xingyi Semiconductor, it is equivalent to completely offended TSMC. In the future, if you want to use TSMC's advanced craftsmanship, it is idiot to say dreams."
"We have to consider this in depth."
"You think too much. Even if the 20nm crafts on TSMC next year will not turn in Huawei. Even Qualcomm can't get on. Next year, TSMC's 20nm process is very limited, all of which are given Apple A8."
"Next year is not the turn of us. But the next year, there is hope in the next year."
"In the next year? How do you know that Xingyi Technology cannot achieve a 20nm technology in the next year?"
"It makes sense. With the speed of research and development of Xingyi Technology, it is completely hopeful to achieve 20 nanometer technology in the next year, so you don’t need to care about this."
"The most important thing is that TSMC's position is not firm. This time, for Qualcomm, it directly destroys the contract and breaks the order of Xingyi Technology's 40 -nanometer.
"We continue TSMC's foundry, and we can't guarantee whether they will also break our chips, especially in the future, our mobile phone sales will increase sharply and surpass Apple, then everything is possible."
"Yes, the encounter of Xingyi Semiconductor today is the future of our Hisilicon Semiconductor. We have to consider this in depth."
"TSMC's position has always been unstable. They used a lot of old rice technology and had to obey each other. And Xingyi Technology was completely different. They were all self -developed technology. Everything said it."
"Ke Xingyi mobile phone and Huawei are competitive relationships. Do you rest assured that you hand over the order to the competitors? Don't you be afraid of being stuck?"
Yu Dong waved his hand: "Don't worry about this, President Wang promised, no matter what happened, Xingyi Semiconductor will continue to give Huawei foundry chips. TSMC dares not promise and cannot be guaranteed."
"And Wang Dong's personality will not break our chips. I personally think that Xingyi Technology can be trusted."
"Speaking well, when the critical time, what can be reliable is brother business!"
The hearty voice sounded, and Ren Lao walked in from outside the door:
"Although we have competition in mobile phone business with Xingyi Technology, it can be a benign competition, common progress, and common development."
"In fact, now Huawei mobile phones, other domestic mobile phones, and even global mobile phones are learning Xingyi Technology and imitating Xingyi mobile phones.
"As for the problem of disconnection, don't worry. If Xingyi Technology is offset, unless they don't want domestic sales."
Yu Dong nodded again and again: "Yes, if Xingyi Technology breaks the Chinese chip, the Chinese people can drowns them one by one. Don't want the sales of Xingyi mobile phones."
Ren Laoxiao said: "And Wang Dong can't do such a thing, don't worry about this. But it is very likely that TSMC's discipline.
Yi Semiconductor. "
"Subjected!"
"Subjected!"
Everyone nodded, and there were no other opinions.
As Yu Dong and Ren Lao said, Xingyi Semiconductor will not be disconnected like TSMC. This is the biggest strategic significance.
As for suppressing Huawei mobile phones, maliciously disconnected ... Unless Xingyi mobile phone does not want to sell it in China.
This kind of triggering things will not happen.
The next day, Zhang Chao came to Huawei Headquarters again.
Lao Yu warmly received: "Mr. Zhang, welcome to welcome, know you are coming, I push all arrangements, prepare tea, wait for you."
Zhang Chao's heart is settled. It seems that the foundry of the foundry is stable. The key is how to talk about the details!
"Thank you Mr. Yu for your enthusiastic hospitality, this time I also came with sincerity."
"Sincerity, then I can look forward to, hahaha." Yu Dong said with a smile: "Really not concealed, we have also discussed internally about Xingyi Technology's foundry. Some people support and some people oppose it, but in general
, Still support much. As for the foundry price, I don't know ... "
Zhang Chao did not answer directly, and asked, "How much does Kirin 910 be mass -produced?"
"20 million pieces!" Yu Dong said decisively.
Sure enough, as the chairman expected!
Zhang Chao continued: "There are other chips next year? 20 million can not be enough Huawei's appetite."
This year, the annual sales of Huawei mobile phones are around 140 million units, and the proportion of smartphones is more than half, about 70,000 to 80 million units.
Of course, function phones also require chips, such as Qualcomm's dual -core, Snapdragon 200 series, Mergotent, dual -core, entry -level quad -core, TEGRA 2, etc.
processor.
However, the current star conductor is still blank at this entry -level processor.
Even the worst 40 -nanometer 3G chip Kunpeng 500, the actual experience is comparable to the Snapdragon 400, plus the bonus of the 64 -bit system, almost catching up with the Snapdragon 600.
The special chip of the function phone is really not.
The production capacity of the Xingyi wafer plant is limited, and Wang Yi cannot see the entry -level processor, or the function machine processor.
That's right, the entry -level processor such as Snapdragon 200 is the purpose of functional machines.
The first generation of red rice in the previous life is not available, at least it must be the low -end processor of the Snapdragon 400 level.
Yu Dong did not conceal: "Yes, we will also produce Kirin 920 next year, and the Kirin 920's modified Kirin 925, Kirin 928, a total of three chips."
"What is the expected production capacity of these three models?"
Yu Dong understands that Zhang Chao is to quote Huawei according to Huawei's order volume!
This is of course.
The chip foundry industry, the typical scale effect, has always been watching people.
The larger the volume, the lower the price of the foundry, which is taken for granted.
Yu Dong decisively said: "Kirin 920 is the main product, Honor 7 will be used, and the output should be around 10 million. Kirin 920 Huawei Mate 7 is used, and the production capacity is also 10 million.
One, a total of almost 25 million pieces! Coupled with Kirin 910, a total of 45 million pieces! "
Lao Yu is full of confidence, as if we are saying that we are so large, there is no water.
Zhang Chao did not believe it at all, and muttered in his heart: ‘Next year, the total sales of Huawei’ s smart machines will be 780 million, and the Kirin chip model accounts for a small proportion, and it cannot exceed 50%.
‘Say 45 million pieces, remove 30 % of the water, and about 30 million pieces.’
"Kirin 910 Huawei P6S, Mate 2 flagship mobile phones, as well as the Honor Thousand yuan machine, it is almost 20 million."
‘As for the Kirin 920, 3 or four million killed. It’ s a little more, seven million Kirin 925 of Mate 7. ’
‘As for the Extreme Edition Kirin 928, Extreme Edition, what sales are very small, and 5 million?
"In this way, a total of 30 million heads, no more."
Zhang Chao is instantly clear everything and does not disassemble:
"For Huawei, Wang Dong has explained it, and gives special support. It stands to reason that the foundry price of Huawei Kirin 910 is calculated according to the 20 million yuan in Kirin 910.
Metering calculation. ""
Yu Dong nodded, this is the industry market.
Like a 28 -nanometer high -end mobile phone chip, one million yuan, the foundry price will be as high as 150 yuan+.
By 5 million levels, the foundry price will drop to about 120 yuan.
In the tens of millions of levels, the foundry price will drop to about 110 yuan.
By the time of 20 million levels, the price of foundry will drop to about 90 yuan.
If it is 50 million or more, the price is only about 70 yuan.
Typical scale effect.
It is also the reason why small companies cannot afford chips.
If you do less, if one million pieces, the price of the foundry will cost 150 yuan, the diluted research and development cost is also scary, all of which add up, the cost is more than 300, which is more expensive than buying ready, and the performance may not be good.
"
No way, the 28 -nanometer process this year is the most advanced craft, which is expensive.
Wait until 20 nanometers, 16 nanometers are all on, and 28 nanometers will be cheaper.
Zhang Chao Feng turned: "But this is the first time we have cooperated. In addition to your 45 million volume, we can give a discount that has never been before.
910 chip price, come to calculate. "
"Truth?" Yu Dong was a little surprised!
The 5 million volume is the price of 5 million foundry, and the price of 20 million is 20 million foundry prices. The two are completely different!
Like Kirin 910 million yuan in volume, TSMC's telegraph price is 90 yuan.
Subsequent Kirin 920, Kirin 925, and a single volume of 400-7 million pieces, which takes about 120 yuan.
As for the Kirin 928, which holds millions of dollars, the price of foundry is 150+!
Yu Dong has a headache!
Even if Xingyi Technology gives some discounts, it is only a few dollars.
But now, Zhang Chao means that the Kirin 920, Kirin 925, which originally wanted 120 yuan, was based on the price of the Kirin 910. It only costs 90 yuan?
It is equivalent to 30 yuan for the unit price of these two chips, and more than 3 million will be cheaper!
"Is there such a good thing?" Yu Dong was a little incredible.
"Of course, if the total ordering of the three chips in one year reached 45 million, the subsequent Kirin 920/925, all at the ultra -low price of Kirin 910." Zhang Chao said with a smile.
"This ..." Yu Dong was silent, 45 million was blowing, and it was impossible to achieve.
Even if it is Huawei, this year will not be sold hotly.
On the contrary, mobile phones equipped with Kirin chip can be 30 million a year, which is very good.
"If 30 million ..." Yu Dong couldn't help asking.
"Thirty million, that's a bit difficult. After all, the Kunpeng chip purchased by other brands is 30,000 to 40 million, right?" Zhang Chaoyi said deeply.
Yu Dong listened to the meaning of the other party, and muttered in his heart: ‘This is that I want the Kirin chip order to order enough, let me use the Kunpeng chip to make up ...’
‘But it’ s not! ’
‘In the next year, Kirin chip is estimated to account for only 40 % of the smart machines, and the remaining 60 % still have to buy out.’
‘Either buy MediaTek, or buy Qualcomm ...’
‘But you can also buy Xingyi Technology’ s Kunpeng chip, which can also make Kirin 920, Kirin 925 at ultra -low -cost foundry! ’
'TSMC thinks that we have a small volume, and the price of the foundry given is not cheap. As for the price of Kirin 920 and Kirin 925, the price is calculated at 20 million yuan.
. '
‘The price of Xingyi Semiconductor is much more interested than TSMC, and it’ s not impossible to use Kunpeng chip. ’
‘And Kunpeng chip is 64 -bit, strong performance, and 64 -bit improvement, which can increase sales!’
Yu Dong quickly decided: "This is okay. As long as the foundry price is low, we can purchase 40 million Kunpeng chips.
"Of course." Zhang Chao got up and reached out: "Looking forward to our long -term strategic cooperation."
"I also look forward to it." Yu Dong said with a smile.
Subsequently, Yu Dong quickly asked for Dong Dong to discuss this.
Ren Dongxi's eyebrows: "This kind of cooperation is very good. Whose chip we buy is to buy. Buying Xingyi Technology chips can also reduce the price of foundry. Why not?"
"Then let's purchase 40 million Kunpeng chips?"
"Yes, and the price of the foundry of Kirin 920 and Kirin 925 is low, and you can also produce more points."
"I understand. In addition, in the follow -up film, I also asked Xingyi Semiconductor to give us priority, so long -term cooperation is fine."
Yu Dong made up a word, TSMC has a great cause, and the streaming films have to be queued.
They are all apples, Qualcomm, MediaTek, Nvidia, Hisilicon ...
That's right, Nvidia is in front of Haisi.
As for MediaTek, the volume is much larger than Hisilicon, Hisilicon can only be lined up.
For example, next year's 20 -nanometer process, Qualcomm did not rank, all gave Apple.
The next day, Yu Dong saw Zhang Chao again. After a simple consultation, the two smiled:
"President Zhang, happy cooperation!"
"Happy cooperation!"
The two sides officially reached a strategic cooperation, which is in line with the interests of both parties.
As for Xingyi Semiconductor OEM, it will leak the feet of Huawei's chip ...
This is not the case. After all, Huawei's current chips are 32 -bit, and Xingyi Semiconductor is at all. The starting to start is 64 -bit processors. How can the 32 -bit processor in the old -fashioned 32 -bit processor reverse the project?
This is just a joke.
Just like a doctoral student, to copy the composition of a elementary school student, it doesn't make sense.
And Xingyi Technology's chips are more powerful and more advanced. Even if they are 64 -bit processors in the future, they don't look at it.
Don't worry at all.
Two days later, the two sides officially signed a contract.
Xingyi Semiconductor gave Huawei unable to refuse the price, follow -up Kirin 920, Kirin 925, all calculated the price according to the volume of Kirin 910, all 90 yuan!
As a exchange, Huawei purchased 40 million Kunpeng chips in Xingyi Semiconductor at one time.
Among them, 5 million Kunpeng 706 SOC, 10 million Kunpeng 508 SOC, 5 million Kunpeng 508C, 20 million Kunpeng 505 SOC.
And in terms of ultra -low OEM prices, Huawei not only placed 20 million Kirin 910 at one time, but even the order volume of Kirin 920 increased to 8 million pieces. The order volume of Kirin 925 increased to 10 million pieces!
If 120 Kirin 920,925, Huawei is not willing to buy more.
But 90 Kirin 920,925, do not take the opportunity to place more orders, are all fools.
Coupled with 1 million Kirin 928, a total of 39 million flagship chip foundries are placed.
Of course, 1 million Kirin 928 must be at a normal price of 150 yuan.
The order amount is too small. If it is 90 yuan, Wang Yi will lose money.
Buy 40 million Kunpeng chips, in exchange for the ultra -low foundry price of two flagship chips.
Huawei feels worthy of more than 400 million foundry fees.
The Xingyi Semiconductor grabs a big customer with infinite potential, anti -killing Taiwan's electricity, and selling 40 million 28 nanometer Kunpeng chips at one time, and won 39 million foundry large orders!
Wang Yi feels very cost -effective, and obviously earns more.
For a big customer, there is nothing to make less foundry fees.
Moreover, the selling Kunpeng chip has returned again.
Like Kirin 920, even if the volume of eight million, according to the price of 90 yuan, Xingyi Semiconductor also earns a lot.
Without him, the cost of 28 nanometer flagship chips, the cost of Xingyi Semiconductor is about 50 yuan.
Even if 90 yuan foundry, there are 40 yuan profit.
A chip profit is 40 yuan, and 39 million pieces are more than 1.56 billion.
As for the 40 million chips sold, although the low -end chip is the main profit, a single profit is above 150.
40 million pieces are 6 billion+!
Give Huawei's 450 million foundry fee discount, replace it back to 1.56 billion foundry profits and 6 billion chip profits.
450 million becomes 7.56 billion, which is completely cost -effective!
It can only be said that this is a very happy cooperation between Xingyi Technology and Huawei. Only TSMC injured cooperation!
However, both sides have no publicity.
From negotiations to signing contracts, they are all low -key.
But there is no windy wall in the world.
Within a few days, TSMC got the news.
"Xingyi Semiconductor has grabbed Huawei's order?"
Lao Zhang Zhang's big mouth, a bit back, but he could not come back: "Aren't they lack of production capacity? How can they still grab our customers of TSMC?"
"Chairman, Xingyi Semiconductor lacks 40 nanometer capacity, but there is no shortage of 28 nanometer capacity. So Huawei's orders have been taken away."
"Made!" Lao Zhang was angry: "Wang Yi, this guy, is to get revenge on our backbone!"
"It must be, we promised Qualcomm to discontinue the 40 -nanometer chip of Xingyi Technology.
"But it is no wonder Xingyi Technology. Instead, we were suddenly broken off, resulting in heavy losses. We will be angry and retaliate."
"That's what he said, but we can't swallow this tone."
"Huawei orders are not as good as Apple and Qualcomm, but it is also a level of more than 30 million, especially the Kirin 910. They place an order at one time and 20 million. Now they have been snatched by Xingyi Technology. I can't swallow this tone."
"You can't swallow the order of 20 million chips a year? Let's give up the order of Xingyi Technology, but 20 million a month! How can you swallow it?"
"That's not that there is no way. It is more important than Xingyi Technology, obviously Qualcomm, right?"
"Even if we have the fault, but Xingyi Technology grabbing our TSMC, we can't just finish it."
"Okay, what do you say?"
"Let's give 10 million pieces for the last 30 million chips produced by Xingyi Technology, and there are 10 million unprepared codes?"
"What do you mean, drag, don't give it?"
"That must not work, 20 million chips, although it is 40 nanometer low -end chip Kunpeng 500 and Kunpeng 506, a 25 yuan foundry fee, 20 million pieces also have 500 million yuan.
Want? "
"Not only the payment of 500 million, if it is not delivered, it will pay 20%more liquidated damages, which is 100 million."
"We deduct their chips, which can indeed allow Xingyi Technology to pour fuel on fire and lose heavy losses, but we will pay 600 million yuan."
"Do you think it is appropriate?"
"A breath, 600 million! Is it worth it?"
"If it is 60 million, 600 million, then really ..."
"this……"
Everyone was silent.
At this time, Great Smart Jiang Hong said: "The account is not the case. We deducted 20 million Kunpeng chips and did lose 600 million, but we got 20 million Kunpeng chips! We can take these 20 million units
The Kunpeng chip is sold out! "
"???"
"!!!"
For a while, everyone was shocked, and looked complicated to Jiang Hong.
Is this method a bit too a bit dog jumping the wall?
Jiang Hong continued: "You think, Xingyi Technology's Kunpeng 500,150 yuan a one. Kunpeng 506,180 yuan a one. We pack and sell, all are counted as 150 yuan, 20 million pieces are 3 billion!"
"Losses 600 million, earn 3 billion yuan, net profit exceeds 2.4 billion!"
"You just say it is not cost -effective!?"
Everyone collectively speechless: "..."
Sure enough, the high -level technical route rises, the technology is a good hand, but the overall view is a bit unqualified.
Chapter completed!