Once Xingyi Semiconductor grows, major first-tier cities will undoubtedly go all out to attract it.
They will not only offer factories and land, just like the capital city, but also various preferential policies.
There's no other way; building a wafer fab costs too much, currently requiring 20 billion.
Four hundred billion, or even five hundred billion in a few years!
If various cities rush to offer land, factories, and preferential policies, and Wang Yi is only responsible for providing equipment and technology, then the pressure will be much smaller.
However, the premise is that Xingyi Semiconductor can grow stronger and mass-produce advanced processes.
At the very least, it must surpass SMIC!
Resources in first-tier cities are also limited and will be preferentially invested in leading domestic enterprises to make them stronger and compete with TSMC, Intel, and Samsung Semiconductor.
They will not invest in the domestic perennial runner-up, making the perennial runner-up catch up with the domestic leader.
After all, there's little point in the runner-up catching up with the domestic leader, as it would still be far from international giants.
Therefore, only by surpassing SMIC and becoming the domestic semiconductor leader will Xingyi Semiconductor receive the most resource support, enabling better development to catch up with the three giants!
As for SMIC, which would be surpassed, it would most likely be relegated to a low-end 40-nanometer foundry.
It would only be able to mass-produce 28-nanometer chips many years later, and nothing more.
A group of people who only know how to contend for power and influence lack the ability to defy fate!
Especially with the emergence of the more outstanding Xingyi Semiconductor, SMIC's strategic significance and status in its previous life will be replaced by Xingyi Semiconductor, thus significantly declining!
Director Qin didn't say these things, but Wang Yi saw right through them.
Since SMIC was a lost cause, Wang Yi decided to replace it and go all out to build Xingyi Semiconductor, focusing on high-end products.
As for low-end 40-nanometer orders, they could be given to SMIC in the future, which would also be considered support for SMIC.
After all, Xingyi Semiconductor's own wafer fab capacity is limited; once the 28-nanometer process is successfully mass-produced, it will prioritize the production of advanced 28-nanometer and higher processes in the future!
There's no need for low-end 40-nanometer products to occupy advanced 12-inch production lines; it would be better to outsource them to SMIC.
As for mass-producing 40-nanometer chips now, it's only because 28-nanometer still needs one to two years to achieve a breakthrough.
Wang Yi had planned all of this.
It was already seven in the evening when he returned to Xingyi Technology Park.
"Chairman, you're back?"
Elder Hu waited eagerly.
Wang Yi was a little surprised:
"You haven't left work yet?"
"Hehe, it's no bother. Overtime is normal in the semiconductor department,"
Elder Hu said with a smile.
Not knowing the outcome of the negotiations, he had no mind to go home.
Wang Yi smiled faintly:
"Fortunately, I didn't fail my mission. By working overtime and rushing the construction, the first fab building will be delivered next month. Although it's not large, with a monthly capacity of only 5,000 wafers, it's sufficient!"
"5,000 wafers? That's already very good, perfectly suited to be a research and development fab!"
Elder Hu was overjoyed.
Wang Yi showed Elder Hu the wafer fab information he brought back from Director Qin:
"The entire wafer fab is divided into multiple sections, mainly three major fab buildings. The first small building, with a monthly capacity of 5,000 wafers, will be delivered in April if we rush the construction."
"The second large building, with a monthly capacity of 10,000 wafers, should be delivered in August."
"The third large building, also with a monthly capacity of 10,000 wafers, will be in November!"
"Good, no problem!" Elder Hu's eyes lit up, "With the first small fab building delivered in April, we'll bring in equipment, and debugging will take three to four months. This way, we can start trial production in July or August. If successful, we'll be able to mass-produce 40-nanometer chips by the end of the year. If trial production fails, we'll adjust and still be able to mass-produce 40-nanometer chips in the first half of next year!"
"Very good!"
Wang Yi was in high spirits.
Forty-nanometer technology had already been self-developed, and self-production promised a bright future.
"Elder Hu, I'll leave these materials with you. Here are also the contact details for the chief engineer and relevant persons in charge of the wafer fab construction. Please arrange for someone to contact them to discuss next month's handover and equipment installation."
Elder Hu's face lit up with joy, and he solemnly took the materials:
"Don't worry, Chairman Wang, I will personally lead people there tomorrow!"
Although it was only Xingyi Semiconductor's first wafer fab, starting with the most advanced 12-inch technology, Elder Hu felt it was already quite remarkable.
"Okay!"
Wang Yi nodded.
If all goes well, this wafer fab will fully mass-produce 40-nanometer chips next year, with a monthly capacity of 25,000 wafers.
Low-end chips like the C1 fast-charging chip are relatively simple; one wafer can yield over a thousand C1 chips.
A monthly capacity of 25,000 wafers means 25 million C1 chips.
That's 300 million chips a year.
By then, there will be 100 million each of P1 power management chips, C1 fast-charging chips, and H1 smart home chips.
If capacity is insufficient, TSMC can continue to be used for foundry services.
As for the second wafer fab, it will be delivered around the end of 2013 or early 2014.
Once the 28-nanometer process is successfully developed, it will be perfectly used for mass-producing 28-nanometer mobile phone chips.
28-nanometer mobile phone chips are more high-end and complex, with many times more transistors, so fewer chips will be yielded, estimated at only 600-800 pieces per wafer.
Wang Yi remembered that the Kirin 990 in his previous life, at 7 nanometers, could only yield about 640 chips per wafer.
And factoring in the yield rate, removing defective chips, there would only be over 500 pieces left!
However, once 40-nanometer mass production is successful, Xingyi Semiconductor will have caught up with SMIC and firmly established itself.
By then, Wang Yi will have more say, and the second wafer fab provided by the capital city can also be built larger, with more capacity.
For example, achieving a monthly capacity of 30,000 wafers, or even 40,000!
This way, it can also produce over 300 million 28-nanometer high-end chips annually.
Xingyi mobile phones, Xingyi tablets, and Xingyi TVs, all combined, will have enough, or even more than enough, capacity, allowing some to be allocated to other products.
By then, one 12-inch 40-nanometer wafer fab for low-end chips and one 12-inch 28-nanometer wafer fab for high-end chips will temporarily be sufficient.
Once 22-nanometer mass production is successful the year after next, the 40-nanometer wafer fab can be fully upgraded to the 22-nanometer process, and the outdated 40-nanometer chip orders can be handed over to SMIC.
Wang Yi had meticulously planned all of this.
Next was waiting!
Waiting for April, for the first fab building to be delivered!
Waiting for August, for 40-nanometer trial production!
Waiting for the end of the year, for the entire 12-inch wafer fab to be fully delivered!
Waiting for next year, for the 40-nanometer process to see soaring yield rates and full mass production!
The future looked promising!
Wang Yi was in high spirits, but then he thought of an important question:
"Elder Hu, the 40-nanometer process isn't very difficult for us, right?"
Elder Hu waved his hand:
"Not very difficult. By the end of this year, we guarantee successful mass production of 40-nanometer chips! At that time, we'll simultaneously start 28-nanometer R&D to catch up with the semiconductor giants!"
"Is 28-nanometer R&D very difficult?"
Wang Yi asked meaningfully.
Elder Hu nodded:
"Developing and mass-producing 40-nanometer chips within a year is completely stress-free. But 28-nanometer is one of the most advanced processes currently; R&D alone will take a year, and mass production at least a year and a half."
Wang Yi had an idea:
"In that case, can we recruit more people, split them into two teams—one continuing to develop the 40-nanometer process, and the other starting to develop the 28-nanometer process now? This way, we might hope to mass-produce 28-nanometer chips by the end of next year!"
Wang Yi knew that TSMC always had two R&D teams, simultaneously advancing two generations of process technology.
For instance, currently, one team is working on 22-nanometer.
And another team is already working on 20-nanometer!
This is even more true for Xingyi Semiconductor, which started late.
"This..." Elder Hu thought for a moment, "This is feasible, but it requires more manpower and a larger investment!"
Wang Yi waved his hand:
"Money is not an issue. We are currently behind, so we need to use capital to buy time! Go ahead and do it without restraint; I'll handle the funding. If you need anything, directly contact Jian Ziyan; all of Xingyi Technology will fully cooperate with Xingyi Semiconductor!"
"Good, good!" Elder Hu was in high spirits, "Since Chairman Wang said so, I'll split into two teams to tackle 40-nanometer and 28-nanometer simultaneously. We will strive for 40-nanometer mass production by the end of this year, and 28-nanometer mass production by the end of next year!"
"Chairman Wang, I'll go prepare right away!"
"Thank you for your trouble!"
Looking at Elder Hu's slightly stooped back, Wang Yi's expression was somewhat complex.
Elder Hu was born in 1947, making him 65 years old this year!
But there was no other way; Xingyi Semiconductor needed to develop and needed brilliant minds, so Elder Hu would have to work until his last breath...
In fact, Elder Hu and Elder Liang were very similar; both came from a technical and research background, were not good at deceiving subordinates or flattering superiors, and wouldn't speak unless they were confident.
Elder Jiang, however, was different; he was better at maneuvering for power and influence, and was highly favored by Morris Chang...
But Wang Yi didn't think highly of such people; he still admired practical types like Elder Hu and Elder Liang!
Elder Hu said that 40-nanometer mass production would be by the end of 2012, and 28-nanometer by the end of 2013, which meant it was almost certainly true!
By then, in 2014, Xingyi mobile phones would be able to use self-developed and self-produced 28-nanometer processors.
However, by 2014, Qualcomm's processors would already be at 20-nanometer, still two generations ahead of Xingyi Semiconductor.
But thinking that Qualcomm's first 20-nanometer processor was the Snapdragon 810, Wang Yi felt no pressure.
The infamous Snapdragon 810, it was so disastrous, it nearly wiped out several half-dead international brands in one go!
"With Elder Hu, Xingyi Semiconductor has a promising future!"
Wang Yi was in a great mood.
Next year, Xingyi mobile phones, Xingyi tablets, and Xingyi TVs will all be able to use self-developed and self-produced processors.
The next day, Elder Hu personally took people to the wafer fab.
Wang Yi, meanwhile, sat in his office and began meticulously planning for the budget phone!
With the budget phone's price set, costs must be compressed; it's impossible to use a 300-yuan Qualcomm processor.
In his previous life, in 2013, the Redmi 1st generation developed by Lei Bus failed.
It wasn't until February 2014 that Xiaomi successfully released the Redmi 1S, which was also the first successful Redmi phone.
It featured MediaTek's low-end quad-core processor and Qualcomm's low-end quad-core Snapdragon 400 processor; both used 28-nanometer process technology and were low-end SoCs with integrated basebands.
And Wang Yi's budget phone would most likely be released in mid-2013, more than half a year earlier than the Redmi in his previous life.
However, to control costs and support the development of the domestic industrial chain, budget phones mostly use domestically produced components.
But currently, domestic component suppliers are just starting and developing, and many areas are not yet mature.
The Redmi 1st generation in his previous life failed because some components were inadequate, resulting in a loss of 50 million in R&D funds.
Similarly, Wang Yi's budget phone would also face similar problems.
The industrial chain, from scratch, requires a development process; rushing it will only lead to complete failure.
Therefore, the R&D cycle for the first budget phone was almost catching up with that of flagship phones.
Various issues in the domestic component industrial chain must all be resolved one by one.
Otherwise, they would all follow in the footsteps of the Redmi 1st generation.
Furthermore, Wang Yi would not entrust the development of the budget phone to Zhu Changlin's department.
They develop Xingyi mobile phones, focusing only on high-end products.
The budget phone, however, would be an entirely new brand, located in Jizhou, and a completely new department would be built for it.
Two product lines, two positions, completely different.
Xingyi mobile phones could be high-cost, high-quality, and highly innovative, aiming to provide consumers with the best phones, even regardless of cost.
For budget phones, the first rule was cost control!
Simply put, Xingyi mobile phones could use whatever was best, while budget phones would use whatever was cheapest yet didn't affect normal use!
These were unavoidable circumstances.
It seemed that when he returned to Jizhou in a couple of days, he would have to discuss the budget phone with Song Sining.
The budget phone would be released in mid-next year, so recruitment, R&D, and advancing industrial chain negotiations must start now.
As for the processor choice, it was also a big problem.
Next year's Qualcomm mobile phone processors would all be SoCs with integrated basebands, and their prices would be the total price of the processor + baseband.
For example, next year's high-end Qualcomm chip, the Snapdragon 800, would cost around 300.
Compared to Qualcomm flagship processors costing over a thousand yuan years later, this price was considered cheap.
Of course, most domestic phones currently only cost 2,000 yuan, not 4,000 yuan like they would years later.
As for next year's mid-range Snapdragon 600 processor, it would also cost around 200!
And the low-end Snapdragon 400 processor, used for budget phones, would also cost around 150.
Nvidia's processors were cheaper, but they lacked basebands, and with the added cost of a baseband, they would even be more expensive than Qualcomm 400.
As for MediaTek processors, they would be a bit cheaper than Qualcomm's, likely available for just over 100.
The material cost + marketing and amortized R&D costs for current flagship phones must all be kept under 1,500.
As for the material cost of budget phones, it must be controlled at around 500 yuan!
For example, the Redmi 1S in his previous life had a material cost of 520 yuan.
Adding amortized R&D costs, taxes, marketing costs, and after-sales costs, selling for 799, it could earn about 200 yuan per unit.
As for claims of a mere 5% net profit, that was just fooling fools.
Even the 1999 Xiaomi 1st generation had a profit of around 500, a profit margin of about 25%!
The same applied to budget phones.
Xingyi Technology's budget phones also needed their material costs controlled at around 500.
Adding taxes, amortized R&D costs, marketing, after-sales, etc., the total cost would be nearly 600, selling for 999, yielding a profit of 400 yuan per unit.
Although this couldn't compare to the profit of Xingyi flagship phones, sales would definitely explode.
In 2013, when smartphones all started at 2,000 yuan, Wang Yi's 1,000-yuan smartphone had no rivals; it was cheaper than many feature phones and knock-offs.
As long as capacity was sufficient, it could sell 100 million units a year!
Earning 400 per unit, that's 40 billion a year!
Small profits but quick turnover, it was that powerful.
Of course, the most important aspect was the underlying user base and market share capture!
For budget phones, with costs to be controlled around 500 yuan, the screen alone would be 130-150 yuan.
After all, even for a budget phone, next year it would need a 720P high-resolution screen; even AUO or Tianma screens would cost at least around 130.
For the processor module, MediaTek 105, Qualcomm 150!
Memory + storage chips, using the cheapest 1GB RAM + 8GB ROM, would also cost around 80.
Casing + metal buttons, 50 yuan.
Casing + plastic buttons, 40 yuan.
The remaining camera module, touch module, battery, etc., 270.
Calculating this way, using the lowest-cost options for everything, the total cost would be at least 625.
Completely over budget!
There was no other way; in 2013, the domestic mobile phone industrial chain was not developed enough, and costs were 10-20% higher than in 2014.
Therefore, while the Redmi 1S could achieve a material cost of 520 yuan in 2014, Wang Yi would have to spend 625 yuan in 2013!
To further reduce costs, he would have to switch to self-developed low-end chips.
The material cost of chips was very low.
The key was high R&D costs and high foundry production line costs!
But as long as sales went up, the amortized R&D costs and production line costs would also be very low.
If sales reached hundreds of millions, all amortized costs could be extremely low, and with foundry fees and material costs, it would only be a few tens of yuan.
The current cost-based pricing strategy in the semiconductor industry is mostly 8:20.
That is, if the cost is 8, the price is 20 yuan.
MediaTek, for example, operates this way.
For powerful semiconductor giants like Qualcomm and Intel, it's not 8:20, but 8:32!
For instance, Qualcomm's quad-core flagship SoC, the Snapdragon 800, which will be released next year, costs around 300, but the cost per chip is only about 75 yuan.
The low-end Snapdragon 400 processor sells for 150, with a cost per chip of at most 38 yuan.
Similarly, if Wang Yi self-develops a low-end chip, as long as sales reach 50 million, the cost can also drop to around 40.
If sales exceed 100 million, the cost can drop to around 30!
For 2013's budget smartphones, with sufficient production capacity, annual sales exceeding 100 million would not be an issue.
Qualcomm chips 150 yuan, MediaTek 105 yuan, self-developed 30 yuan!
In this way, the total cost of the budget phone could be further reduced to 550 yuan!
Selling it for 999 then would yield a profit of 450 yuan per unit, which would be quite substantial.
As for the R&D of budget phone chips, that would also be easy; it could be directly derived by downgrading this year's Kunpeng 700.
By reducing the clock speed and downgrading some specifications, it would become the Kunpeng 500, which could then be taped out together with the Kunpeng 700.
Alternatively, the low-end Kunpeng 500 could be taped out first; with the successful experience from the 500, there would be greater confidence in the Kunpeng 700.
However, there were also two R&D options for the budget phone's low-end chip.
One option was 28-nanometer, outsourced to TSMC for manufacturing, costing 30 yuan.
The other was 40-nanometer, self-produced at the Xingyi Wafer Fab by the end of the year, which would result in even lower costs.
Forty-nanometer was already considerably cheaper than 28-nanometer, and being self-produced, it saved on foundry fees, making the estimated total cost only 20 yuan!
Moreover, the Snapdragon 400 would only be released in the fourth quarter of next year, and phones featuring the Snapdragon 400 would come out the year after next.
The Kunpeng 500 would be launched early next year, a year ahead of schedule, so even the 40-nanometer process would be sufficient!
At worst, slightly less reduction in clock speed, and its performance might not be worse than the Snapdragon 400!
After all, this year's Nvidia quad-core Tegra 3 processors are all 40-nanometer, so Wang Yi's budget phone using the Kunpeng 500 next year, also a 40-nanometer quad-core processor, would be sufficient!
After all, for a budget phone, what else could one ask for? Cost was the top priority!
This was completely feasible.
Forty-nanometer mass production would be achieved by the end of the year, and by then, the Kunpeng 500's tape-out could be handled in-house.
As for having too many chips and finding R&D difficult?
That wasn't an issue.
After all, the Kunpeng 700, Kunpeng 500, and next year's Kunpeng 900 were actually the same set of chips, with the same architecture.
The only differences were in clock speed, and whether the baseband was external or integrated.
In short, the Kunpeng 700 was the standard version with an external baseband, the 500 was the downgraded version with an external baseband, and the 900 was the upgraded version with an integrated baseband!
They were all derived from a single set of chips!
Furthermore, with the successful trial production of 40-nanometer chips in the second half of the year, the Kunpeng 500's tape-out wouldn't need TSMC anymore; Xingyi Wafer Fab could handle it directly!
That would make it even easier!
Wang Yi made up his mind that with high sales projected for budget phones next year, the chips would have to be self-developed and self-produced.
With this thought, Wang Yi once again went to the chip department to find Williams.
"William, I have some good news for you."
"What good news?"
Williams became interested.
Wang Yi smiled faintly:
"Our first wafer fab will have its first building delivered next month. Elder Hu and his team will immediately bring in equipment to use it as an R&D factory, aiming for 40-nanometer process trial production in August. If all goes well, we can mass-produce 40-nanometer chips by the end of the year!"
Williams' face lit up with joy:
"Mass production of 40-nanometer chips by the end of the year? With Elder Hu's capability, there's absolutely no problem! This truly is good news, excellent!"
Wang Yi then changed the topic:
"There's another piece of good news."
"What is it?"
Williams became interested.
Wang Yi said with a smile:
"Next year, we will launch a budget smartphone, with costs controlled around 500, priced at 999, and aiming for 100 million annual sales!"
"Oh my god!" Williams exclaimed in disbelief, "What kind of divine idea is this? It's incredible."
"Five hundred cost, budget smartphone, hundreds of millions in sales!"
Williams couldn't help but stand up:
"Chairman Wang, wait, let me think about this!"
"If successful, that would be 40 billion in profit!"
"Crazy, truly crazy!"
"And it's practically feasible!"
"With this phone, Xingyi mobile phone sales next year will exceed 150 million! Sweeping the globe!"
"By then, the user base of Xingyi mobile phones will reach an incredibly massive scale."
"The significance behind this is enormous!"
Williams instantly thought of many things.
A budget phone, with a profit of only 400 per unit; he clearly knew that Wang Yi was definitely not doing it just to earn that 400 yuan.
Instead, it was for hundreds of millions of smartphone users!
Wang Yi smiled knowingly:
"To keep the cost at 500 yuan, we'll need a 40-nanometer low-end quad-core processor!"
Williams' eyes lit up, and he instantly understood Wang Yi's meaning:
"Chairman, that's easy. It's just a low-end processor, no need for extremely powerful performance, very simple."
Saying this, Williams pulled out a proposal:
"Chairman, this is the Kunpeng 700 we've currently designed, but it's not yet perfect and still needs further adjustments."
"But if we cut down some features, lower all clock speeds, and switch to a 40-nanometer process, using it as a low-end processor wouldn't be a problem."
Wang Yi nodded:
"That's what I meant too. We'll directly make two versions of the mobile chip: a 28-nanometer flagship Kunpeng 700 chip, and then a downgraded 40-nanometer low-end Kunpeng 500 chip. Once the 40-nanometer process trial production is successful in the second half of the year, we can proceed with the Kunpeng 500's tape-out!"
"No problem!" Williams responded decisively, "With the experience from Kunpeng 500, we can better develop the Kunpeng 700, and even the Kunpeng 900!"
(End of chapter)