Rewriting the Technological Scenario

After AMD gave the wind semiconductor order and guarantee the yield, the news that AMD officially cooperated with the wind group.

Kil Singer came to China again, attended the cooperation conference with Meng Qian, and announced the depth partnership between the two companies at the press conference.

This cooperation will involve multiple fields, including computer chips, including mobile phone chips, including artificial intelligence, including service platforms.

In cooperation, this high-wind semiconductor guarantees AMD's recent production only a gift from the wind group, and the wind group truly brings AMD, it is more important.

That is: amdyes!

A chip industry has a well-known infarction.

The AMD's fire reborn three major guidelines after Dr. Su, creating great products, deepening partnerships and simplified operations.

And these three are actually achieved a goal, the product is preferred.

The chip market is very important, but the core is still looking at the product, especially for the rice companies, it is basically not worried that it is pressed by his own country, and it is most important to do product quality.

Dr. Su is apparent to this, so the product is placed first, and Dr. Su, who knows the AMD weakness, puts the bet on the desktop, and it is true that she is pressed.

Now, Kilsinger chose the wind group for AMD.

Great wind group, 3D transistor from the world, although only more than one month, but the morning is also a brother, and a brother is also a minute.

Great Breene Group, one of the world's two major photolithography leading companies, one of the world's three major chip IDM companies, one of the world's four major chip architecture companies, one of the world's five major chip patents.

This cooperation, the wind group will first help AMD solve two problems with compatibility and power consumption.

AMD's compatibility has not been Intel. This is also something that there is no way. Many games and software will be adapted to the Intel platform and the Ying Weida graphics card. It is more focused on the CPU, but there is no chipset without the graphics card, and the compatibility problem is more obvious.

But now the big wind graphics card is pressing the Yingweida, the wind group has its own ecology, and AMD will come to a platform to solve compatibility.

AMD power consumption is also old problems, this is exactly a dominant technology of wind semiconductors, and there is no unexpected words to see obvious optimization within one or two years.

However, let Kilgeg look forward to another thing, it is because of the comprehensiveness of the wind group, although in the design and manufacturing of the computer chip, there is still obvious distance from Intel, but the future development blueprint of the wind semiconductor is very concerned. .

As a partner, Killa Singer can see the Semiconductor Development Bluepruetors of the Great Breeze Group in the next decade. Of course, the details don't think, after all, Kilzeg is clearly clear, the wind group will do the high-end computer chip soon, look back Everyone is also an opponent, but it doesn't have a absolute two choices for AMD. It is also gambling.

The semiconductor development blueprint for the wind semiconductor next decade is actually finalized in the beginning of the year.

It is well known that Intel starts to jump from 14nm, pitted a bunch of customers, Samsung Terrane electricity began to engage in 5nm crafts, Intel is still with 10nm, and even Intel will handle 7 nm orders to Taixue Electricity.

Then many people said that Intel did not fall. Intel technology can't, in fact, this is not so simple to evaluate, this is a bit complicated.

If you want to say from density and performance, Intel's 10nm ++ is definitely playing with 7nm. It is not said that 7nm must be better than 10nm. Without this statement, the key to this matter is the semiconductor development of each company before and after 2010. There is a difference in blueprints.

How to develop the chip's future chip, will have five years and ten years of development blueprints, rather than saying that this year is thinking next year, this year is already five years later, and even 10 years later.

In 2011, the chip started to enter 14nm attack. At this time, there was a problem, and then push it forward, it is 10nm, and the process should be forwarded. In the current technical theory, EUV is definitely a best choice, but EUV This thing doesn't know how many years I have to know.

And the EUV lightning machine is not that it can be used directly. One of the conversion processes in the middle is very complicated. The EUV lightning machine is more important. It is also only one ring in the entire chip manufacturing. After using the EUV optical machine The overall technology of supporting, this is generally possible to mature two or three years, which leads to the difference between the planning of different enterprises.

Samsung is the most powerful one, earliest Samsung's EUV program, Intel obviously be the most conservative in this issue, and repeatedly stated that EUV's maturity should be 2020 years later, Intel uses EUV to be at least 2020 After that, the 7nm process is also available.

TSMC is in the middle, and finally it is still a lonely, and it has become the king in this field.

So in Intel asks if Intel's 10nm does not need EUV, in fact, the reason is very pure, because the plan is settled in the plan.

However, Intel may have lacking a little luck, Intel's most conservative, a very important reason is that it is really burning, Intel can completely scrubbing DUV, research.

On the contrary, the technical backward tables and Samsung must take it. According to the normal market forega, Samsung and TSME can't see this matter in the short term of EUV. From 2016 to 2017, the Thai electricity is coming three times. It is possible to see that the TSM does not use EUV to work 10 nm.

The problem is that Samsung and Terminal Electric goes with Intel Gang DUV technology, just.

So the point of standing in that era is not wrong. It may be wrong to invest too many new technologies in Intel in 14nm and 10nm processes. In order to preempt the technology, it is too radical, and it has met Typical electricity. Magical adapts to digestive EUV technology in 2 years.

The reason why TSMC can be preserved earlier than the completion of EUV, but also benefit from smart mobile devices.

TSMP is a pure foundry factory. The rapid development of smart mobile devices has brought huge amount of substitute experience, good product rate, and leakage rate, etc. in the foundry of TSMC. Experience and great effort. In the process, it is constantly improving, the advantage is too big.

This natural advantage makes the Taiwan Powering When it is steadily stabilized, the development and cost of the advanced process and the cost of the company have arrived in the corporate, so there is only the sentence, the success of TSM is high-pass Apple and Huawei.

So the fact is actually proved that Intel's pre-judgment is true, because Intel does not have Apple Qualcomm Huawei to provide him with a cost-owned cost and provide a very large amount of experience.

If the Intel's blueprint is 10nm, the Intel's probability will only be more cool.

Samsung is also aware of this, so announced the chip foundry independently in 2017.

Intel does pre-judgment accurate, and it is indeed that the technology of the 10nm crafts of Intel's DUV is not bad. It can even say that the world may have only Intel can use DUV to make this 10nm process, this is Intel. Absolute strength.

It can be used by intelligent mobile devices to use the smart mobile device to use the EUV technology. This may be Intel's life.

Your technical cattle can not stand the transformation of the times, smart mobile devices, and disqualify the monopoly of Intel established in again and again.

However, after 2020, EUV has entered Intel's plan, and I will really change anything after I start using EUV.

Therefore, the wind semiconductor is now developing in this 10 years of semiconductor development blueprints. This program will have a great impact on the development of the future.

From the current technological development situation, the probability should be achieved before 2015, so although it is not possible to determine it, Meng Qian still decides that 10 nm begins to use EUV.

After making this decision, in the process of continuing to attack the 14nm, the company can start experimenting through an immature EUV prototype, if there is a small probability accident, the influence will be large, the future is in the computer I want to fight with Intel on the chip.

The wind semiconductor wants to be very difficult to do with Intel on DUV. This is not a money to solve, can only hope that the fate will not deliberately .

At the same time, another important thing is to develop the foundry of the wind semiconductor. This is why Meng Qian is so positive this year to find the reasons for the extension order.

According to the ten-year development blueprint of the wind semiconductor, the company's foundry industry will go out independently. The next plan is 14nm before 2014. It is expected to be EUV lightning machine in 2015. EUV light before 2017. The momentum mass production is 10 nm, and 2018 challenges 7NM, and becomes a leading enterprise in the 5nm process before 2020.

...