Currently, the largest seller of 3G mobile phones is Xingyi Technology.
Except for the newly released xphone 3 and xphone 3 plus, all other models of Xingyi Technology are 3G mobile phones.
The largest seller of 3G chips is also Xingyi Technology!
Except for Kunpeng 902 SOC, other chips are also 3G chips.
Reports suddenly focused on domestic 4G being commercialized soon, allowing users to choose 4G mobile phones and manufacturers to choose 4G chips...
If this wasn’t for Xingyi Technology, Wang Yi wouldn’t believe it!
It's already a conspiracy.
"We are already investigating, but it will take some time!" As soon as this incident came out, Zhu Changlin arranged for people to take action.
Wang Yi nodded: "But with the atmosphere heightened to this point, it seems that 4G is on the verge of launching. 4G chips must also be fully deployed. Let's go and have a look at the chip department."
Regardless of whether he came for Xingyi Technology or not, Wang Yi had to prepare for the worst and be prepared.
This is his usual style.
Later, Wang Yi and Zhu Changlin came to the chip research and development department together.
"Chairman, you are here." George greeted him with a smile.
"How is the 28nm 4G chip upgrade plan going?" Wang Yi got straight to the point.
As a reborn person, Wang Yi knew that the 4G era in China was about to begin, and it was impossible for Wang Yi to be unprepared.
On the contrary, the 28nm 4G upgrade plan started a few months ago!
"It's all done. Kunpeng 706 SOC, the 28nm 4G upgraded version of Kunpeng 702, and Kunpeng 706C have been successfully taped out and have begun mass production."
"The 28nm 4G upgraded version of Kunpeng 506, Kunpeng 508 SOC, and Kunpeng 508C have also been successfully taped out. Since they are temporarily unavailable, mass production has not yet begun."
"In addition, we also upgraded the entry-level Kunpeng 500 SOC to 28nm 4G and made it into Kunpeng 505 SOC and Kunpeng 505 C! Together with several other chips, they have been successfully taped out and can be mass-produced at any time.
.”
Wang Yi understood George's idea, which was to map out Qualcomm's four chip series one by one!
This idea is very good, very good.
Mobile phones are divided into flagship phones, mid-range phones, low-end phones, and entry-level phones. The chips are divided accordingly and completely matched.
"How is the performance after the upgrade?" Wang Yi asked the key, hoping not to disappoint himself.
"It's very powerful. Even if you don't count the 60% improvement of the XOS system, it can outperform Qualcomm's chips of the same level, or even surpass them!" George was full of confidence:
"In the field of flagship chips, the paper performance of Kunpeng 902 SOC is 25% better than that of Snapdragon 800. Coupled with the 60% improvement of the XOS system, it is 85% better!"
"The upgraded Kunpeng 706 SOC is 25% stronger than Kunpeng 702 and about 35% stronger than Snapdragon 600. With the 60% improvement of the XOS system, it is 95% stronger!"
“The Kunpeng 706 SOC is 35% better than the Snapdragon 600? Doesn’t that mean that it is almost catching up with the Snapdragon 800, not counting the 60% bonus of the XOS system?”
Wang Yi was shocked. Snapdragon 800 is only 40% better than Snapdragon 600!
This is a significant improvement across generations.
As for the Snapdragon 801AA, it is only 10% stronger than the Snapdragon 800. The Snapdragon 801AB is 15% stronger. The Snapdragon 801AC is about 20% stronger.
On the contrary, Snapdragon 805 has improved a lot, and its theoretical performance has been greatly improved compared to Snapdragon 800, but its actual performance is hard to describe...
George nodded: "That's right, the paper performance of Kunpeng 706 is slightly 5% worse than that of Snapdragon 800, but our tuning is better than Qualcomm's, and we can definitely overtake it. It can be said that Kunpeng 706 SOC can crush Snapdragon 800, and then With a 60% improvement, it can crush the next generation Snapdragon 805!"
"Okay, very good, very good!" Wang Yi was shocked.
Unexpectedly, Kunpeng 702 has been upgraded to 28nm process. After the main frequency is increased, the improvement is so big!
It can only be said that George's team is very knowledgeable and worthy of being the veteran team of Texas Instruments.
The core teams of George's team and Zhang Chao's team are all elites recruited from Texas Instruments.
Then they added the domestic and foreign semiconductor elites recruited later. The overall strength is several levels stronger than HiSilicon, and not much worse than the Qualcomm team.
It can only be said that although Texas Instruments gave up the mobile chip business, the former king is still the king, the king that Qualcomm looks up to.
After all, the Texas Instruments team has always used the 512M main frequency to compete with Qualcomm's 1G main frequency, and it can still maintain its lead. Qualcomm has been tortured to the point of losing its temper and can only pile up parameters like crazy.
Essentially, these R&D personnel from Texas Instruments are very powerful.
Wang Yi felt like he had found a treasure.
When Xingyi Technology started its business, it was Wang Yi's wisest choice to spend huge sums of money to acquire the core chip R&D elites from Texas Instruments!
The second wise choice was to introduce Mr. Liang’s mentor Mr. Hu and his team to create Xingyi Semiconductor and develop 40nm and 28nm processes.
If these two steps are not taken correctly, Kunpeng chip will still be a PPT...
George continued: "Kunpeng 508 SOC is 25% higher than Kunpeng 506, 50%+ stronger than Snapdragon 400, and about 20% weaker than Snapdragon 600. But with the 60% improvement of the XOS system, it is 40% stronger than Snapdragon 600 !”
Snapdragon 800 is 40% better than Snapdragon 600, but Snapdragon 600 is 80% better than Snapdragon 400!
On the contrary, Snapdragon 400 is 20% better than Snapdragon 200.
To put it bluntly, Snapdragon 400 is the most expensive generation and not many brands use it.
In the previous life, it was not until next year's Snapdragon 410 performance was improved that it was used by many brands such as Redmi 2, Huawei, Honor, and ZTE.
"Kunpeng 505 is 26% better than Kunpeng 500, and its performance is close to that of the previous Kunpeng 506. It is 15% better than Snapdragon 400. Coupled with the improvement of the XOS system, it is 75% better!"
"So, our worst 4G chip Kunpeng 505, not counting the 60% bonus of XOS, is better than Snapdragon 400. There is indeed no comparable product to Snapdragon 200. They are too weak. The 40nm Kunpeng 500 is
Much better than Snapdragon 200."
Wang Yi was dumbfounded, George was in Versailles!
The reason is simple. Qualcomm's high-end chips are doing well, and its mid-range chips are derived from last year's flagship chip APQ 8064, which is also very good.
But Qualcomm's low-end chips and entry-level chips are very poor, even worse than MediaTek.
Without him, Qualcomm had never made low-end chips before. The Snapdragon 200 and Snapdragon 400 series were both made in Qualcomm’s first year, so they were naturally very poor.
Next year, Qualcomm will start making second-generation low-end chips and entry-level chips. With experience, it will be much better.
As for why Xingyi Technology’s low-end chips are so powerful?
The reason is simple. Kunpeng 500, Kunpeng 506, and Kunpeng 700 are all castrated from the Kunpeng 702. They have the same CPU architecture as the Kunpeng 702, but the main frequency is different. There is a gap, so they are all successful.
As for Kunpeng 505, Kunpeng 508, and Kunpeng 706, they are just 28nm 4G upgraded versions of the above-mentioned chips, which are naturally stronger.
"Very good, George, you have done a great job in upgrading the 40nm 3G chip, very well!" Wang Yi affirmed and praised.
It is important to develop new chips, and it is also important to get these upgrades.
With these chips, Xingyi Semiconductor has enriched everything from entry-level chips to flagship chips.
And it has dual process coverage of 40nm and 28nm!
3G chips are all made using 40nm technology.
4G chips are all made using 28nm technology.
"Chairman, how will the next chip production arrangements be carried out?" George asked: "Currently, only Kunpeng 902 SOC and Kunpeng 902C are in production, as well as Kunpeng 706 SOC and Kunpeng 706C."
Previously, the production capacity of 28-nanometer chips was limited, and only two production lines of the second and third phases of the Imperial B1 wafer fab were in production.
A total of 50,000 wafers, with a monthly production of 30 million chips.
A total of 45 million Kunpeng 902 SOCs were produced in July and August for xphone 3 and xphone 3 plus.
15 million Kunpeng 902C, used in Xingyi Tablet 2 pro and Xingyi TV 4K version.
In September, 15 million Kunpeng 902 SOCs were produced for xphone 3 and xphone 3 plus.
5 million Kunpeng 706C for Xingyi Tablet 2.
10 million Kunpeng 706 SOC for Unbounded Note.
But other types of chips have not yet begun mass production.
Wang Yi asked: "The first phase of Xingyi B2 wafer fab was delivered more than two months ago. Will it be put into production next month?"
"Yes, Chairman, 28nm chips will be put into production in early October, with a monthly production capacity of 50,000 wafers and a monthly production of 30 million chips."
George replied: "Moreover, the first phase of the Modu M2 wafer fab was delivered half a month ago and can be put into production in early November. It will also produce 30 million chips per month."
"This is great!" Wang Yi was overjoyed: "In early October, we added 30 million new production capacity. Adding the previous 30 million, we can produce a total of 60 million chips per month starting from October. Starting from November, we can produce 90 million chips per month.
!”
Wang Yi fell into deep thought and began to plan carefully.
Kunpeng 902C is used in Xingyi tablet 2pro and Xingyi TV 4K version. 15 million units per month is enough.
Tablets and TVs are not mobile phones. It would be great to sell a few million units a month, but not more than 10 million units.
Kunpeng 706C is used in the standard version of Xingyi Tablet 2. 5 million units have been produced. The subsequent 5 million units per month will be enough. The main sales force is Xingyi Tablet 2pro.
Kunpeng 706 SOC is used in Unbounded notes, Unbounded high-end products, focusing on cost-effectiveness and high performance. 10 million units have been produced, and the subsequent 10 million units per month will almost maintain a sales balance.
However, as a 4G sub-flagship chip, it has to be sold to third-party customers, and an additional 5 million chips will be purchased first.
As for the Kunpeng 508 SOC, not counting the 60% bonus of the XOS system, it is 50% stronger than the Snapdragon 400 and 20% weaker than the Snapdragon 600. It would be a bit of a loss to compete with the Snapdragon 400.
Even the Kunpeng 505 is better than the Snapdragon 400 and can catch up with next year's Snapdragon 412.
It would be better to use Kunpeng 505 to compete with Snapdragon 400, sell it to third parties, grab Snapdragon 400 customers, and simultaneously make a 4G version of the unbounded mobile phone, which will be launched together next year, and 5 million units will be produced every month.
As for the Kunpeng 508 SOC, it is almost catching up with the Snapdragon 600. It will be reserved for next year’s Unbounded 2 generation and there is no rush to produce it for the time being.
Compared with Snapdragon 400 and 200, Kunpeng 505 SOC is enough!
The remaining production capacity will all be used to produce Kunpeng 902 SOC.
It will also be delivered in batches to the top ten customers, with a large order of 80 million pieces!
After considering these, Wang Yi spoke directly. George had already taken out a pen and paper and started recording:
"Starting from October, we will produce 20 million Kunpeng 902 SOC units per month."
"Monthly production of 15 million Kunpeng 706 SOC."
"Production of 5 million Kunpeng 505 SOC per month."
“Monthly production of Kunpeng 902C is 15 million pieces.”
"Monthly production of 5 million Kunpeng 706C units."
"In November, production will continue according to the original plan. The new 30 million capacity of the M2 wafer fab will be planned according to the chip order volume in October. The remaining part will be used to generate Kunpeng 902 SOC!"
The most critical thing is Kunpeng 902 SOC, after all, there is a big order of 80 million.
A total of 60 million units were produced in July, August and September for xphone 3 and xphone 3 plus. By the end of September, 32 million units will have been used, leaving a balance of 28 million units.
The expected sales of xphone 3 and xphone 3 plus in October are estimated to be around 20 million units, with a maximum of 25 million units, and a maximum consumption of 25 million chips.
In October, 20 million pieces of 902 SOC were produced, plus 28 million pieces in stock, 25 million pieces were used, leaving a balance of 23 million pieces, which can be enough to deliver 15 million pieces to third-party manufacturers, with a balance of 8 million pieces, to prevent a burst of orders.
The third-party manufacturer placed the order in September, and Wang Yi started delivery in October, which is already very fast.
In November, 30 million pieces of production capacity were added, and 15 million pieces can be allocated to Kunpeng 706 and Kunpeng 505 to cope with third-party orders.
The remaining 10 million pieces of production capacity are used to produce Kunpeng 902 SOC. Subsequently, 30 million Kunpeng 902 SOCs will be produced every month. In addition to 15 million pieces being delivered to third-party customers, the remaining 15 million pieces and 8 million in inventory are enough for self-use.
After all, subsequent mobile phone sales continue to decline.
Such a large order of 80 million chips will be delivered in about 6 months!
Of course, this is just a rough plan, and it will have to be dynamically adjusted based on the number of third-party orders.
George looked at the notes: "This arrangement is just right. Not counting the 64-bit improvement, Kunpeng 505 is 20% better than Snapdragon 400. It is enough to beat Snapdragon 400!"
Wang Yi nodded: "Use Kunpeng 902 to beat Snapdragon 800, use Kunpeng 706 SOC to beat Snapdragon 600, use Kunpeng 505 to beat Snapdragon 400 and 200, even if we don't count the bonus of the XOS system, we are 20% ahead of Qualcomm at the same level -
35%, almost iterative. Counting the 64-bit bonus, it’s a flash sale across two generations. We must kill Qualcomm and grab third-party customers!”
"Chip Marketing Department, take quick action to sell our Kunpeng 902, Kunpeng 706, and Kunpeng 505 SOC as soon as possible. We also have Kunpeng 902C, Kunpeng 706C, and Kunpeng 505C. As long as there are customers who want them, we can produce them!"
"These chips must be sold well, we have no retreat!"
"There are also 40-nanometer 3G chips. As long as there are customers who want them, they will all be on sale!"
There is no way, starting from November, Xingyi Technology’s 28nm monthly production capacity will be as high as 90 million chips!
The monthly production capacity of 40 nanometer chips is 30 million chips.
Together, the monthly production capacity is 120 million chips!
With such a large volume, it is not sold to third-party customers. If Xingyi Technology uses it for itself, it will not be used at all, and the inventory will be ruined!
The second phase of B2 and M2 wafer fabs will be delivered next year, and the Shenzhen wafer fab will be delivered the year after that...
Without more production and less sales, there is no future.
It seems that the OEM matter will also be put on the agenda next year!
"Okay, Chairman, I will arrange the chip marketing department to further contact the major manufacturers, especially the old customers we have cooperated with, and hope to win."
George said with a smile: "After all, the previous Kunpeng 702 and Kunpeng 700 were deeply loved by customers and have accumulated an excellent reputation. They also have chip debugging services, which major customers cannot refuse."
Wang Yi nodded: "Chip tuning service is a technical problem that most mobile phone brands cannot handle. Even Qualcomm and Samsung don't understand it. This is our core competitiveness and golden brand."
Zhu Changlin was also very emotional: "Yes, the actual performance of a chip depends not only on the parameters, but also on the tuning. For example, Kunpeng 508
SOC was originally intended to benchmark Snapdragon 400, but it was too powerful. On paper, its performance was only 20% worse than Snapdragon 600. But after our tuning, the actual performance is estimated to be only 5% worse than Snapdragon 800. If you encounter a useless brand
Negatively optimized Snapdragon 800 may not be as good as our Kunpeng 508!”
"Haha, this is very possible. After all, negative optimization has always been the specialty of major mobile phone manufacturers." Wang Yixiao said.
"How to determine the price of Kunpeng 508?" George asked: "Should it be based on Snapdragon 400 or Snapdragon 600?"
"Kunpeng 508, the performance benchmark of Snapdragon 600, is priced according to Snapdragon 600, which can be slightly lower. If customers find it expensive, low-end chips will buy cheaper Kunpeng 505, which is enough to crush Snapdragon 400. As for the 508, it cannot be sold.
It doesn’t matter if you go out, leave it to next year’s Unbounded 2 generation.” Wang Yi said:
"Including the training service fee, if the minimum order is 5 million pieces:"
"Kunpeng 902 SOC, 4G chip, sold for 300 yuan."
"Kunpeng 706 SOC, 4G chip, sold for 270 yuan."
"Kunpeng 702 SOC, 3G chip, is on sale for 240 yuan."
"Kunpeng 700 SOC, 3G chip, sold for 210 yuan."
"Kunpeng 508 SOC, 4G chip, sold for 210 yuan."
"Kunpeng 506 SOC, 3G chip, sold for 180 yuan."
"Kunpeng 505 SOC, 4G chip, sold for 180 yuan."
"Kunpeng 500 SOC, 3G chip, sold for 150 yuan."
Even lower is the Snapdragon 200, which has even worse performance. With the baseband included, it’s still 120-140. Xingyi Technology doesn’t have this category.
George looked at the record and nodded: "The price is very competitive, especially for 4G chips. For example, Kunpeng 505,180 yuan, Snapdragon 400 is about the same price, but the performance is not as good as ours, and it is not 64-bit. I am worried that they will cut the price."
!”
"It's okay, they cut prices, and we will follow suit. There is room for price reductions in 3G chip prices. As long as Qualcomm cuts prices, all our 3G chips will drop in price by 30 yuan, so it won't be a big problem." Wang Yixiao said:
"As for 4G chips, there is no need to lower them. Even if Qualcomm cuts the price by 30 yuan, we will not drop it. We have the advantage of 64-bit chips, all integrated with 4G baseband, and the performance is 20%-35% more powerful than them. Even if it is 30 yuan more expensive, it is completely
It can be sold!”
Wang Yi thought clearly that this is the lowest price for 4G chips and it will not change within a year.
If Qualcomm wants to fight a price war to death, then it will cut the price of 3G chips by 30% to a cabbage price and kill Qualcomm.
Even if 4G commercialization starts in China next year, 4G mobile phones will only account for 30% at most, and only about 40% of the world's 4G mobile phones.
On the contrary, most of the remaining mobile phones, especially mid- to low-end mobile phones, are still 3G.
4G packages are expensive, so developed countries promote them quickly.
But more users are still in developing countries, and they all use 3G.
Even in developed countries, many users are still using 3G mobile phones.
Therefore, Wang Yi will not reduce 3G chips to cabbage prices unless absolutely necessary.
But if Qualcomm is ignorant and launches a price war, then Wang Yi will have no choice but to give up the car and save the handsome guy.
Use cabbage-priced 3G chips to kill the opponent!
Just like the dispute between SMIC and TSMC in the previous life, TSMC has a monopoly on high-end technology, but mid- to low-end chips are the big player!
Similarly, in the chip market next year, 3G chips will still account for more than half!
In the following days, the entire Xingyi Semiconductor became busy.
In particular, the B2 wafer fab is entering final preparations to begin mass production of 28-nanometer chips.
Two days later, Zhu Changlin came to Wang Yi: "Chairman, we found out that the news that domestic 4G will be commercialized by the end of the year was broken by overseas media."
"It looks like it's really Qualcomm's work!" Wang Yi instantly made a clear judgment: "Are they planning to use this to suppress the chip sales of our Xingyi Semiconductor, so that Qualcomm's 4G chips will sell well?"
"That should be the case. After all, Qualcomm's 400 series are 4G chips. Both the 200 and 600 can be plugged into 4G basebands and can also be used as 4G mobile phones."
Zhu Changlin said with a smile: "They probably think that our other chips from Xingyi Semiconductor are all 3G chips. They have no idea that we have already achieved a 28nm 4G upgrade! And they have already started mass production."
"Yeah, Qualcomm didn't expect that we had prepared a 4G chip upgrade plan a few months ago!" Wang Yixiao said: "But at this point, the chip war between us and Qualcomm has fully erupted. The other party
Now that we have taken action, we will also show our swords. Inform all major customers!"
"Okay, what day is the time set for?"
Just as Wang Yi was about to speak, there was a knock on the door and Jian Ziyan walked in:
"Chairman, the latest news is that Qualcomm will hold a 4G chip exhibition on October 5. It will not only have high-end Snapdragon 800, mid-range Snapdragon 600, but also low-end Snapdragon 400 and entry-level Snapdragon 200.
There are four series in total, all 4G, all 28nm process, available for sale globally."
"Sure enough, this matter was done by Qualcomm. We first laid the groundwork for 4G and launched a public opinion offensive to let all manufacturers know that China will also launch 4G next summer and must make 4G mobile phones. Then we will hold a 4G chip exhibition and invite major manufacturers.
As a result, major manufacturers will be excited and place orders one after another."
Zhu Changlin eloquently said: "Most orders for high-end chips are won by our Xingyi Technology. Most orders for mid- to low-end chips and entry-level chips are monopolized by Qualcomm. They have a good idea!"
Wang Yi nodded: "It is indeed a good plan, but unfortunately, an accident occurred!"
"The news is that on October 3, we will hold a 4G and 3G chip exhibition, and we will have everything you need!"
"Okay, Chairman." Zhu Changlin was in a good mood.
Just two days in advance, Qualcomm will be killed!
Soon, the two pieces of news shocked the entire Internet, and netizens were delighted.
"Qualcomm just announced the chip exhibition on the 5th, and Xingyi Technology announced the exhibition on the 3rd. The meaning of tit-for-tat is too obvious."
"That's right! Domestic chips have been suppressed and beaten. Now that Xingyi Technology is rising strongly, it's time to show its sword!"
"A bright sword is a bright sword, but Xingyi Technology only has one 4G chip and four 3G chips, which is a bit difficult. Qualcomm has Snapdragon 200, 400, 600, and 800, covering everything from entry-level to high-end, all 4G chips, or plug-in 4G baseband.
"
"Yes, domestic 4G has also started at the end of the year, and Omi 4G has been launched for many years. Xingyi Technology only has one 4G, and it will definitely not be able to handle it. It is a bit self-defeating to have a trade fair on the third day."
"What does it mean to humiliate oneself? Can't we give domestic products more time? Kunpeng 902 is more powerful than Snapdragon 800. It's not like you haven't seen it."
"Yes, Xingyi Technology is powerful in high-end chips, but what about mid- to low-end 4G chips? What about entry-level 4G chips? Xingyi Technology doesn't have them at all!"
"Qualcomm has long popularized 28nm, and has 4G chips in four dimensions: high, middle, low and entry-level. This is the gap!"
"Success in one dimension doesn't mean anything. Xingyi Semiconductor still needs to work harder."
Chapter completed!