With the completion of the adjustment and optimization of the two 32nm process production lines, the Dragon Factory began to have the production capacity of four advanced process processes of 100,000 pieces, and Zhang Rujin finally began to show his edge.

This time he's targeting Apple's processor machining.

"Mr. Jobs, we will support your A5 R&D and tape-out with the best process technology, and I believe that SS will not open up their latest 32nm process technology to A5, they need this valuable process technology to support their high-end memory production. ”

Zhang Rujin has long seen the weakness of his competitors.

"And, most importantly, Mr. Jobs, don't you think that SS's products are extremely homogeneous with yours?

And our dragon factory is different, we will never enter the product field, focus on OEM production for customers, each customer we will have extremely strict confidentiality requirements and capacity commitment guarantee. ”

After laying the groundwork for so long, Zhang Rujin finally released his big move.

Jobs took his pipe in his mouth and quietly listened to Zhang Rujin's words.

"I heard you've mass-produced a new batch of SPARC T2?" Jobs asked a seemingly unrelated question. "Now tell me, what is your yield?"

To the outside world, this data should be extremely confidential trade secrets.

However, Zhang Rujin knew Jobs's intentions, which was worried that the Dragon Factory was the real strength.

"We have started production of three batches, from the first batch of 25 percent, to the third batch of our yield has exceeded 57 percent, if there is a plan for the fourth batch of mass production, I believe our yield will not be less than 75 percent!"

Zhang Rujin said confidently.

As a large order with a top challenge at the beginning, it is not easy for Dragon Factory to finally achieve a yield rate of more than 50%.

Jobs really raised his eyebrows, and he began to have the most intuitive feeling for the Dragon Factory.

Commercially speaking, if it can achieve a yield rate of 75%, then it already has a good prospect of commercial mass production, at least according to this yield, even if the SPARC T2 processor is priced at $500, Dragon Factory has a good profit.

Of course, thanks to the overpricing of that batch of SPARC T2, Dragon Factory made a lot of money on that big order, because most of the production capacity was realized by the last batch.

If it is really produced according to the yield rate of more than 20 percent, the dragon factory can barely keep the capital enough.

But with a technology maniac like Liang Jisong, how can he endure such a low yield rate for mass production?

Therefore, although the number of this order is not much, the dragon factory is still divided into three batches of production according to the plan, and the stop in the middle is naturally to adjust and optimize the performance of the production line, and the yield rate increases more than one time, if it is not for the mainland customers urging the goods, Liang Jisong can also optimize another version to reach more than 75% before large-scale mass production.

"We won't pay for the pre-trial production and tape-out of the foundry, Zhang, I think you should know this, in addition, our processing fee is not higher than that of SS because you are a 32nm process process. ”

Of course, Jobs does not want to be monopolized by the foundry, especially now that the production and sales of fruit machines have increased significantly, and the chip foundry will be a major event for the company in the future.

In fact, TMS has come to the door and asked to take over their business like Zhang Rujin, and Jobs has promised them the opportunity to test the stream.

The conditions are exactly the same as what Jobs said to Zhang Rujin just now.

"No problem. Of course, Zhang Rujin knows the truth of the customer bullying the store, if it is a small customer, to introduce the 32nm process technology, even if it is to pay millions of yuan for the trial flow Zhang Rujin may not agree, after all, this is the core treasure of the dragon factory, and it is a big loss to delay one.

For businesses that have not seen large-scale orders, Zhang Rujin will definitely not waste time in the future.

"Well, I wish us a happy cooperation. Jobs stretched out his hand and shook it tightly with Zhang Rujin.

Jobs grinned with his pipe in his mouth.

Shop around for a good deal.

This Zhang, the same Zhang who came last time, are all good people.

Well, if they all have no problem trying out, maybe you can consider taking turns to produce A5.

To be able to take the Big Apple on the road to make money, Jobs considered something, of course, not just standing on a huge stage and making a strong endorsement for his products!

There are not many big orders in the OEM industry that are secrets, and everyone is watching big customers very closely.

Whether it's SS or TMS, everyone has long expected this customer battle. In fact, for TMS's graphics card processor orders, SS company has not used its brains.

What's more, now the production line of TMS has been deeply optimized according to the manufacturer, whether it is capacity, yield or price, it is actually difficult for SS company to compete with TMS.

To win this order, first of all, the price will inevitably be lowered to make it possible for customers to relax, OEM costs will be reduced, and the initial yield rate will definitely be inferior to competitors, so that the initial introduction is basically a loss state, and it is possible to make money until the yield catches up with TMS.

But how long is the waiting time to catch up, is it a quarter or two quarters? SS is not very sure, not to mention that its core interests are not in the OEM.

His state-of-the-art production line also has an advantage in its memory business.

The current Qimonda is too ferocious.

Not only has a huge production capacity of ordinary memory chips, but also has the capacity reserve of 4 advanced process processes of the Dragon factory, once the fire is exchanged in high-end production, the Dragon Factory plus the Saxony factory, the pressure on the SS company is too great, and he can only pull Hynix together to fight.

Similarly, Zhang Rujin can move the Apple processor A5, which seems to have fallen into the fat of SS company, but he can't move TMS's big cake in terms of graphics card processor, which is also the same reason as SS company.

If the two sides are to exchange fire at the high end, Dragon Factory will inevitably be involved in a brutal price war, and a part of the advanced process must be retained as a backing, and it must output a steady stream of price war ammunition at full power.

At the critical moment, we will output enough high-end storage particles to the market, such as the latest specifications of 4Gbit memory chips, and turn the competitor's niche into a red sea to see who can survive the cold winter of the market, in other words, who is more financially strong and can endure the opponent to death in a bloody war!

In the current OEM battlefield, the trend of dividing the world is becoming more and more obvious, of course, TMS still takes the lead in occupying the best advantage, with the best customers and high-profit OEM projects.

However, the dragon factory is constantly accumulating strength and improving its own process technology, and now the mass production of 32nm process technology has not become the end of the technological progress of the dragon factory, on the contrary, this is just the beginning.

Now Liang Jisong's full research and development of the 28nm process technology and the pre-research of the 22nm process process are in progress, and the 28nm process process has made a considerable breakthrough.

The arrival of Liang Jisong, a semiconductor processing technology maniac, is certainly not just as simple as optimizing and improving yield on the existing production line.

The most important task is to jump the R&D beam of new processes and new processes.

To this end, Zhang Rujin has paid at least ten core high-end talent indicators to introduce talents everywhere.

All of these are core technical backbones with an annual salary of up to $500,000 to $1.5 million, and are top chip processing and design experts recruited from all over the world.

All from IBM, Intel, AMD, Toyo, Baodao industry famous companies, these chip design or process experts have added a lot of confidence to Liang Jisong, otherwise he alone would not be able to hit a few nails.

These experts come from different technical routes and have different ways of thinking, which just provides Liang Jisong with more ideas to solve problems, so that there are few times when he is at a loss when he encounters problems in the promotion.

Liu Ruoyi, who has deep pockets, will not save money at all in terms of hardware environment, after Zhang Rujin made a request, he directly agreed to build two sets of main top equipment of the production line for them to conduct various tests, you must know that these equipment can be priced at 1 billion US dollars at every turn, and the main equipment of a process production line is an idle asset if it cannot be developed, but now the dragon factory is not afraid at all.

If there is a shortage of money, Liu Ruoyi will immediately continue to raise funds and burn money inside.

Under the condition that there is no shortage of external conditions and no weak internal technical strength, the technological progress of the Dragon Factory has maintained a trend of rapid iterative spiral.

Even if they occasionally stop moving, they will soon solve the problem and trot forward.

With this kind of research and development progress, Zhang Rujin's confidence is very sufficient.

The 32nm process technology is now one of the mainstream top configurations in the current generation of industry, of course, the better 28nm process technology competitors have begun small-scale mass production, and in terms of technology generation, there are two generations of differences from the 45nm process technology to the 28nm process technology.

However, in this way, there will be no restrictions on the introduction of 45nm process technology by Zhongxin International in the mainland.

After all, the dragon factory in Xiangjiang is not suitable for large-scale production, and the large-scale mass production of truly ordinary-level semiconductor products is better cost-effective only if it is placed in the mainland.

In terms of the current mainland production capacity of Zhongxin International, the Dream Factory with a monthly output of 400,000 wafer memory capacity that has been integrated is the core, plus the previous 150,000 other 130nm to 90nm process process capacity, and the 100,000 high-end process process capacity of the Dragon Factory, the overall total annual production capacity has reached 7.8 million wafer processing capacity.

Although it is still far from the 11 million wafer processing capacity of the leading TMS, it is much more than the 4.8 million wafer processing capacity of the original second place in the industry.

After breaking through the 32nm process, Zhang Rujin introduced the 45nm process technology for Zhongxin International in the mainland, and the monthly production capacity design is to start a 150,000-piece 12-inch wafer production line.

In addition, in the production of wafer raw materials, Zhongxin International has also begun to make layouts.

Zhang Rujin is preparing to build a large-scale 12-inch wafer production plant for Zhongxin International in the mainland, with a design output of 300,000 pieces per month in the first phase and more than 500,000 pieces in the second phase in the future, which makes Zhongxin International enough to meet most of the wafer needs of its foundry products.

In terms of this critical semiconductor raw material, Liu Ruoyi's principle is that there must be alternatives, and it is better to have a higher cost than become a bottleneck.

So far, Zhongxin International Mainland and Xiangjiang have raised more than 40 billion US dollars.

Most of them are financed by Liu Ruoyi's investment institutions, so that the equity of Zhongxin International is now basically concentrated in Liu Ruoyi's hands.

Although there is no hope of recovering the cost, but from the books, last year's rapid increase in production capacity has contributed nearly 6 billion US dollars in revenue, if you remove large-scale capital expenditure, book profit has been realized, of course, plus large capital expenditure, Zhongxin International is still unable to achieve independent profit development.

This year's revenue is at least $8 billion, if nothing else, and while it is still some way from competitors' estimates of more than $10 billion, the distance has begun to shrink dramatically.

In this OEM super competition, UM has been unable to prepare for large-scale operations in advanced processes.

With their current annual processing capacity of 4.8 million wafers, most of them are above the 65nm process technology, and now the construction of a 45nm process production line is limited to 50,000 pieces of monthly production capacity, which is really tight on capital, and their financing is not as easy as Liu Ruoyi to get money.

Originally, the second millennium became the third, and the decline in the status of the industry slowly lost its attractiveness in the eyes of investors.

If it weren't for the restrictions of the island's policies, Zhang Rujin would have to be iron-hearted to bring UMC over, and with UM's current weak stock market performance, Zhang Rujin could easily be acquired from a commercial point of view alone.

Now UM can even buy the price of a new 32nm process production line of Zhongxin International.

It's a pity that the island will never let this kind of slap in the face happen. UM is different from the stragglers of memory OEMs, which are the core enterprises on the island alongside TMS.

Zhongxin International is hung with the prefix of the Chinese word, whether it is now that the equity has been dispersed to the hands of offshore investment companies, the legend has fallen into the hands of the financial queen of Xiangjiang.

As long as the foundation of this enterprise is in the mainland, under the banner of the mainland, the island will never let it acquire UM, and even TMS will not be able to try it several times, and Zhongxin International will not have this possibility.

Even if Liu Ruoyi buys it through a third party, it is impossible to achieve a merger with Zhongxin International.

Therefore, Zhang Rujin does not have much memory of UM, and is directly preparing to launch a production capacity war, and let UM strike out through the market, maybe he can also receive some UM customers.

Although these customers may be mostly cheaper than competitors.

However, in the process of centralization, Zhongxin International will not be soft, after all, most of the production capacity is now placed in the mainland, in terms of land costs, labor costs and competitors can not be compared.

The land over there is basically a free gift, and even the factory building is built in the local high-tech zone, and the rent is free for 3 years, and the rent is half price for 5 years.

The new 150,000-wafer production capacity plant and land are basically the same treatment, although it is a bit far from the city, but it is promised that a bus route will be opened to facilitate access.

In order to stand out in this industry, the current situation is not blind in the mainland.

The importance of the 45nm process technology is beyond doubt, not to mention that the SPARC T2 was completely open-sourced last year, and although the 45nm process technology is worse than the 32nm process technology, it is also much stronger than the original 65nm process technology.

The efficiency of this processor is clear from the fact that the major computer research and development institutions in the mainland are now secretly fighting for dominance.

Although the confluence of Jiangnan and Southwest has the upper hand, it is not the only voice, and the value contained in it is not only seen by Li Shijie and Zhao Mengguo.

So the real scramble has just begun.

The mass-produced Shenwei SPARC server has opened up the market in the southwest and Jiangnan, but it is difficult to move forward in the local market, even in the Lingnan market, where the local strength is weak, Shenwei's sales have encountered close pressure treatment from competitors, and the price has been targeted and blocked.

After all, the Shenwei server is still a new star, and the original old forces are very close to Shenwei.

This makes Li Shijie and Zhao Mengguo a little helpless, and there is not much differentiation in the market now, after all, the opponent is also using SPARC compatible servers to play Shenwei's SPARC servers, especially relatively mature products with less than 16 channels, everyone has the technical support of the SPARC Open Alliance, and it can be developed.

In addition to the advantages of the 32-channel super minicomputer, Li Shijie and Zhao Mengguo are now at a loss, and now they can only bet on the new generation of 64-channel SPARC super small computer servers and the new generation of SPARC T2 PLUS.

Relying on technology to have new competitors can not copy the product, they can get out of this red sea market war, high-end products to form a technology generation advantage to hit the opponent, low-end through the scale of the formation of a certain cost and price advantage.

But now the two of them have finally seen the future clearly, and they are still full of confidence in the future.