ARM Wu Xiong: Three Development Trends in Semiconductor Industry

As we all know, 2011 is the post-PC era. In March, Apple’s CEO Steve Jobs announced the arrival of the post-PC era. The entire electronics industry, especially the semiconductor industry, saw the post-PC era driven by mobile computing through the mobile Internet, which is very large for the entire industry chain. opportunity. However, it is also a huge challenge to China's semiconductor industry. I want to share with you the opportunities, challenges and trends in the post-PC era.

Before this, I would like to report good news to everyone. In the past five years, ARM's partner in China and China's design and manufacture of core ARM products have increased industry value by three times that of the entire semiconductor industry. We have always said that ARM has helped the global semiconductor industry get a rapid increase. From the United States to Taiwan to help the rise of South Korea's Samsung, we saw in China to help the rise of China's industrial Chinese core. Our actual output value over the past five years has been seen. The average growth rate of global ARM's partners is more than double the semiconductor industry's (1.2-1.3 times), and it truly realizes the idea of ​​helping China's core.

Not only that, in 2011, we saw that China's partners based on ARM shipments from the past 20 million -30 million to the current 500 million, which is a quantum leap. This represents a core improvement in the global competitiveness of our industry's ability to design and mass produce. However, this is just a beginning. We have mentioned that mobile internet has great opportunities for our industry. why? Let us look at some data. Everyone in the world today has an average of 2 network links. In 2011, everyone consumed an average of 3 SoCs. One of them is based on ARM, and 1.5 of these are based on 8-bit and 16-bit ARMCUs. What will happen in the future? According to Cisco's data, the amount of data used by mobile internet users will grow by a factor of 10 in 2014. That is to say, the performance of the past 8-bit and 16-bit is far from meeting the needs of the new generation of mobile Internet. But this is only part of it. What's more important? More importantly, by 2025, each person's link will exceed 100, which means that from the current Internet (whether mobile phone or PC) model, it will be surrounded by the entire network, our daily life, the entire working environment. Will be surrounded by mobile internet links. Driven by this, for the entire industry chain, this opportunity is not only a multiple of SoC, but also an opportunity of several tens of times. But it is not easy for the industry to truly achieve global success. Because the semiconductor industry is different from e-commerce and the Internet, we have to pull it out. We are playing global competitiveness because the semiconductor industry is a global industry. For us, we must understand the trends and challenges of global development. Only in this way can we truly succeed. What kind of trend is that?

I say three trends today. The first is diversity. The post-PC era is driven by the personal application service experience of the Internet and is driven by consumers. It is a completely different industry driven by the scale effect of PCs. In this industry, we see that the diversification of SoC has taken place. If we recall that we were still doing something 13 years ago five or six years ago, when many industry people thought that our ICs and SoCs would be less and less, only 40 companies might do 40 nanometers at 40 nanometers globally, but today There are more than a dozen of 40 nanometers in China. Therefore, the diversification of semiconductor SoCs is obviously an irresistible trend. This trend is driven by the PC era after the mobile Internet. Whether it's from touch screens to tablets, etc., this is the opportunity for new SoCs waiting for us. We only have so many opportunities when we have two links, and the chances of reaching 100 links are great. Everyone should be very clear.

Diversification itself brings another question. What is it? The product's challenges are very large. In the past, the PC may have been replaced by four or five years, and television may have been longer. However, our current mobile phone is changed once a year. More importantly, now whether it is Android or Other, after the chips are more, our greater work is on the software, because we need to support more application services, especially with third parties, with the entire software and hardware ecosystem. The link, the link support of these application services is very heavy for the entire industry. For the SoCs that are present here, the software that has been added over the past few years is still higher than the hardware. All of our partners have invested two to three times as much hardware in the world, including China. Our investment is so much that the chip still sells so much money. What's worse is that the previous product was sold without price cuts for a year or two. Now it may be a price reduction in ten months. This poses a very big challenge for us. This brings a second trend.

The second is the development of platforms. Because of the changes in these industries, our platform for doing SoC can only be done with platformization. Through the platform, we can accumulate the core technologies and IP of hardware and software, and constantly launch new products through this platform. The most obvious is Apple's iPad, iPhone, etc., completely through the same platform and chip, including Samsung's mobile phones, televisions, tablet, including many domestic companies, such as Hass, etc., through the same platform is unstoppable trend. After this point, another issue has emerged, that is, the selection of the platform is a very big challenge. why? If your entire company's lifeblood is based on this platform, you have chosen the wrong platform, which also determines the future fate of your company. There are several very important factors in choosing the platform. The first one is whether the platform has brought applications in the post-PC era. The entire industry has opened up through interconnection and intelligence, which means that the entire industry chain is long. For the simplest example, we are now talking about mobile payments. The payments in the mobile industry chain are far greater than any payment industry chain in the past, from software and hardware to back-office services, etc. In the new post-PC era, the industry The length of the chain determines the platform selection must choose one of the most extensive third-party ecosystem supported platforms. Only in this way can we really succeed.

What is the second alternative? Power consumption problem. Really choose a low power consumption, from the beginning of the terminal to the back of the cloud to achieve a low-power platform. Why do you say that? This is back to the third trend in the post-PC era.

The third is low power consumption to the cloud. In the past, we all talked about low-power terminals, and those who do semiconductors knew that the reduction in power consumption that has been brought to the 28-nanometer technology has become less and less. This has become flat. From the device of the terminal, we see that on the right side of this figure is a basic standard smartphone battery, which is made of 45000Cal, but its capacity is much lower than that of a candy of the same size, so organic chemistry is better than inorganic chemistry. People are still much stronger. This trend cannot be changed. If we consider the current nanotechnology, in fact, by the year 2020, the same size battery will only increase by 2.2 times. In other words, the power consumption problem on the terminal side has always been accompanied, but this is only the beginning. why? There is also a problem with the cloud. Every 110 mobile phones need a high-end server to support it in the cloud. If more than 100 links are constantly running, how many servers are needed to support it? This support is not only a question of the computing power we talk about, but also the actual physical limitations. For the simplest example, the process of reducing power consumption has reached the point where it is the most gentle, that is, it has reached its peak. Then there is no way from the CPU design, CPU design has a limit on the power of each processor, so you can only add more processors. After adding so many processors, how does it still heat on the original machine? The obvious problem is that storage cannot be solved through more integration. So in the cloud of power consumption has been forced to be mentioned by all the industry chain a very important process. As we all know, more than a month ago, HP announced 32-bit servers based on ARM. The cooperation is to reduce the power consumption on the server side. This cooperation is not only supported by HP, but also supports the entire industry, especially the entire ecosystem of hardware and software. In the era of cloud computing, it does not mean that cloud computing can be independent of the entire industry chain. All standards, technologies, and all new things are based on the original architecture. Therefore, we must consider the support of third parties.

Why are third parties willing to support this matter? Let's take a look at how HP sees it. Where is the benefit of using ARM architecture as a server? On the right is a traditional X86 server. When it needs 400 servers, this computing power requires 10 RAPs. But with the same computing power on the right, ARM-based processors integrate 1600 processors, but its size is only 1/20. What is the power consumption? 1/9. The trend just mentioned to the cloud is very obvious. What are these trends? Brings the evolution of the entire CPU or processor industry. As a leading supplier to the CPU industry, we feel very deeply about this. In 1992, the world's largest processor forum had 32 companies with 32 architectures. In 2009, there were only 4 left, and later or not, because so many people did not participate. The entire industry chain software and hardware, especially software, such a long industrial chain, we can invest together to truly support the general processor platform. After the Internet needs so many applications and services, everyone needs to do less and less. What will happen by 2025? We do not know whether the last is X86 or ARM or coexist, we will wait and see.

It will be exactly the 10th anniversary of the establishment of ARM in China next year. Our approach or vision is very simple. We hope to provide this technology platform to Chinese industries and help the rise of China's industry. Just as we helped the rise of the semiconductor industry in the United States, it is like helping the rise of the semiconductor industry in Taiwan, just as we have helped the rise of the semiconductor industry in Korea. The work of the past few years proves that we are doing faster and better in China than in Taiwan and South Korea. Of the more than 900 industries in our world, 250 in China are the fastest growing.

Looking ahead, we see that the future of ARM, the future of the semiconductor industry, especially the future of China's semiconductor industry, is in your hands. Our business model is to provide our platform, and ultimately help you achieve the final product and achieve the final result. Success. The past few years have proven our success with this industry and we have achieved tremendous growth. We see that in the next decade, in the post-PC era, we will have greater success and growth with China's semiconductor industry. The future is in your hands.

LED Flat PAR Light

Longdi Light Equipment Technology Co., Ltd , http://www.gz-ledlight.com