CMOS ups and downs: from fighting between two heroes to made in China

CMOS ups and downs: from fighting between two heroes to made in China

In a sense, the modern mobile phone war has become a camera war.

From single shot to 4 shots, from 8 million pixels to 100 million pixels, today without revolutionary changes, the self-portrait at night can also automatically beautify the camera effect has become the main battlefield for the differentiation of various brands.

Here, Huawei, Xiaomi and many other mobile phone companies are in full swing, but fundamentally, the performance of mobile phones’ camera performance is only superficial, and the mobile phone manufacturers’ scores and publicity are just superficial efforts, which really determine the quality of a mobile phone’s camera function. Make the chip of CMOS image sensor .

The so-called image sensor can convert the light signal captured by the lens into a digital signal that the machine can understand. That is to draw the real world into the world of the chip. Generally speaking, the larger the sensor area, the larger the canvas area and the more details that can be painted. Its performance directly affects the quality of the camera, which is equivalent to the “eyes” of the camera.

This chip is not large in size, but it accounts for 52% of the cost value of the entire camera module. The camera module, which is often a mobile phone, is the third largest source of cost after SoC chips and screens.

At present, in addition to Sony, which can roll up Apple, Huawei, and Xiaomi, which can produce this “canvas”, there are also Samsung and OmniVision , a total of three major players.

The history of global mobile phone image sensors can be summarized as the struggle between Sony and Howe, and the conclusion of Samsung’s profit.

In the process, technological changes and changes in production capacity have successively pushed different companies to the throne; Howe has also changed from an American company to a Chinese company during this period, leading China’s CMOS to start another game from scratch, shocking The war of the throne.

U.S.-Japan technology battle

From the 1970s to the 1990s, the global image sensor market was monopolized by Japan for more than 20 years.

At that time, the world’s image sensor products were dominated by CCD (Charge Coupled Device Image Sensor) devices with high pixels and high manufacturing difficulty . Sony and Panasonic were the best among them.

Relying on its leading position in the field of CCD and storage, Japan once occupied more than half of the world’s semiconductor production. Among the world’s top 10 semiconductor companies, Japan also once occupied 6 exclusively.

As the Americans who invented the CCD, they were naturally unwilling to leave the market behind. In the more than 20 years of Japan’s dominance, American companies have been looking for an alternative technology that can achieve high-quality CCD imaging. At the same time, the production difficulty and cost of this technology are low enough for American companies to defeat traditional CCDs. Down.

In 1995, several Chinese students from China founded a company called Howe in Silicon Valley, which brought hope to the United States.

This young company has been trying to replace the traditional CCD with the traditional CMOS process (complementary metal oxide semiconductor image sensor) in the image sensor.

CMOS ups and downs: from fighting between two heroes to made in China

The difference between CCD and CMOS

If the two processes are compared intuitively, under the current process conditions, the imaging quality of CMOS is slightly inferior to that of CCD, but it can greatly reduce power consumption and cost, and is more suitable for use in the ascendant consumer electronics market at that time. At that time, Howe participated in the famous Comdex computer exhibition in the industry, and after being purchased by many Taiwanese customers, it achieved mass production in just one month [1] .

Since then, CMOS technology has gradually become the mainstream of the image sensor market. The unparalleled Howe began to drive on the fast lane, and successfully listed on the Nasdaq in 2000, cut into the Apple mobile phone in 2007, and became the earliest Apple industrial chain company.

But the CMOS war has just begun.

Sony up, Howe down

At the same time that Howe was listed on the Nasdaq in 2000, the “Old Master” Sony finally put aside its CCD and turned to the CMOS market.

In sharp contrast with the dominance of the CCD market that year, after ten years of transformation, until 2010, Sony had only 7% of the CMOS market, while the rookie Howe’s market share has reached as high as 50%.

No one thought that the change would happen so quickly: In October 2011, Apple launched the iPhone4S. Before the press conference, some media broke the news that Howe will no longer “link the village” as the main rear camera of Apple and will be replaced by Sony.

At that time, many people thought it was just a Fake News that came out of nowhere, but they never thought that on the day of the press conference, Apple proved the authenticity of the message with actual actions. For the latest iPhone4S, Apple will greatly reduce the proportion of Howe’s supply, and instead support Sony to become Apple’s CMOS supplier.

It’s not that Apple is always abandoning it, but because of the skyrocketing demand of tens of thousands of people robbing Apple, Howe’s technology is leading, but its production capacity has long been unable to keep up with Apple’s pace. This involves two classic models of semiconductor production, IDM and Fabless.

1. Sony adopts the IDM model: that is, from manufacturing, design, packaging and testing, the entire industry chain is handled by itself. The advantage of this model is that it has strong production capacity and can fully meet customer needs, but the disadvantage is that the company has heavy assets and huge investment costs.

2. Howe adopts the Fabless model: The company is only responsible for design and R&D, but the production and packaging and testing links are outsourced to a third party. The advantage of this approach is that the investment is small, the huge capital and labor costs in the production process can be reduced, and the funds can be concentrated for technological upgrades, but the disadvantage is that there is a risk of production capacity.

Sony, which has a self-built factory, has made no secret of its ambitions for expansion: In 2011, Sony once invested 400 and 100 billion yen in research and development of CMOS in the Kumamoto and Nagasaki R&D centers to cope with the surge in market demand.

On the other hand, Haowei, although operating with light assets, faced a sudden surge in demand, and Haowei without its own fab, naturally can only silently suffer a dumb loss, falling from one supply to two.

Later, in 2013, it developed a more advanced stacked CMOS than Howe. Taking advantage of Howe’s technology catch-up, workers were allowed to work overtime to produce, which further stabilized Apple’s status as the first supplier of CMOS.

On the Howe side, although Apple introduced the rear cameras of iPhone4S and iPhone5 in 2011, Sony CMOS has been adopted, but in order to preserve the love, the front camera still uses the Howe solution. After Sony has formally formed a double counterattack of technology + production capacity, starting with iPhone 6, all models of iPhone, regardless of the front and rear cameras, all adopt Sony CMOS, and Howe has since been completely kicked out of the fruit chain.

CMOS ups and downs: from fighting between two heroes to made in China

After 2011, Howe and Sony’s CMOS performance are up and down

The industrial model has created different endings for Sony and Howe. After 2012, Howe’s market share fell from the boss to the third with only 11%. Sony completely took over the customers in the high-end market before Howe, with a market share of more than 40%. While Samsung took advantage of the struggle between Sony and Howe, it relied on the support of the terminal and the latest ISOCELL technology to tear away 20% of the market share from the sites of the two giants.

At this point, the pattern of CMOS dual power has been set, and it seems that there is nothing wrong with Howe, but the story is not over yet.

TSMC, Howe, and Sony’s watershed

In 2013, Howe could not beat Sony in the high-end market, and began to move to China to be the low-end market. By 2019, it was acquired by the Chinese semiconductor company Weir. According to the development of the story of defeat and acquisition, Howe is likely to decline, but why not only is Howe not dead now, but it is even possible to catch up with Sony?

This is related to the nature of CMOS image sensors. CMOS is a special product. Everyone knows that traditional digital chips such as CPU and GPU rely on advanced manufacturing processes, while analog chips such as radio frequency and power management chips rely more on the experience accumulation of masters. But CMOS is a “double-headed eagle”: it is an integration of analog and digital circuits.

On the analog side, it needs to perceive subtle changes in light and turn light signals into electrical signals. However, a single pixel is in the micron level, so the analog chip does not require a particularly advanced manufacturing process.

But at the same time, CMOS also needs to convert the electrical signal into a digital signal that the mobile phone SoC can understand. This part is realized by the second half of the digital chip integrated on the CMOS.

Since the key to imaging lies in the first half, and the optimal number of mobile phone pixels has been stopped at 12 million pixels for a long time, Sony’s own factory operation can not only handle it easily, but also accumulate a lot of industry experience, both in cost and technology. .

Until recent years, sensor manufacturers headed by Samsung have adopted the “small pixel” technical route, pursuing smaller pixel sizes and higher pixel levels, copying Sony’s back road with the cost-effective route.

Under this trend, Samsung and Howe have successively achieved pixels below 0.8 micrometers, so that the number of pixels of the same size sensor has risen sharply. Previously, only 20 million or 40 million sensors could be achieved, and it was pulled to 48 million. , Even 64 million or 100 million pixels, this brings consumers a sense of satisfaction: a mobile phone bought for 3,000 yuan, just looking at the parameters, can be several times higher than Apple’s pixels.

CMOS ups and downs: from fighting between two heroes to made in China

(Even Sony uses small pixels to increase the number of pixels while maintaining the same CMOS area)

This puts Sony in a dilemma. Let’s follow it. I have been practicing craftsmanship with old equipment. To make small pixels, I have to replace it with new equipment with advanced manufacturing process. A lot of my own experience has been erased. Don’t follow it, and the main price is new. The mobile phone market in Belgium began to be eroded.

As for Howe, using TSMC’s advanced technology, it quickly entered the market below 3,000 yuan, which has a huge demand for small pixels, and began to break Sony’s copper and iron walls.

On the one hand, the technological route is evolving, on the other hand, the pattern of CMOS and even the entire semiconductor industry is also changing. For example, the birth of TSMC has made it possible for the semiconductor industry to have a vertical division of labor. Sony’s IDM model had advantages 10 years ago. However, as TSMC’s production capacity and process levels have become more advanced, Sony’s factories have begun to fail to catch up with TSMC.

As a result, TSMC has become a watershed for the development of Howe and Sony, which can be analyzed in two time periods:

1. 2010 period: CMOS mainly competes for production capacity and design technology. At this stage, Howe adopts the Fabless model and transfers production to TSMC. Sony adopted the IDM model. At that time, the CMOS foundry level of the Sony factory was stronger than that of TSMC. In terms of design, Sony CMOS products were also more advanced than Howe, so Sony completely overwhelmed Howe during this period.

2. 2020 period: At this stage, TSMC not only becomes the industry’s first in wafer production capacity and process, but also has the 40~28nm CMOS process required for small pixels.

Due to the strong demand for chips, Sony has been unable to meet customer capacity requirements in the past two years, and its own 55nm process performance has gradually reached a bottleneck, and Sony’s self-built process production barriers have begun to weaken. So at the end of 2019, Sony placed orders for high-end chips to supplement TSMC’s production capacity for the first time, and it could also introduce TSMC’s 28nm process.

In other words, in the CMOS production link, Howe and Sony started to stand on the same starting line, which provided the possibility for Howe to catch up with Sony. Moreover, Howe, which has been declining for a long time, also encountered a new opportunity in 2019: The Weir Group from mainland China acquired Howe with 15.3 billion funds this year. Weir has invested heavily in research and development after acquiring Howe, breaking through 48 million and 64 million pixel technologies in 2019 and 2020 respectively, and gradually catching up with Sony’s Samsung in terms of technology, and Howe has hope of returning to the first echelon.

But these must be based on a premise. Howe, which has always been a low-end product in China, will still use it for high-end products now? this is the key of the problem.

“Made in China” Dongfeng

Soon, Howe ushered in a major opportunity for “human use”.

In May 2019, the US government included Huawei on the physical list and prohibited companies such as Intel and Qualcomm from supplying Huawei. Huawei launched the “Spare Tire Plan” and began to consciously cultivate domestic enterprises, accelerate suppliers, and promote the localization process.

According to the Japanese investigative agency Fomalhaut Techno Solutions’ analysis of Huawei’s Mate30 mobile phone disassembly and analysis in September 2019, in the four months after the U.S. ban, in terms of monetary value, the utilization rate of China’s domestic components has risen from 25% to 42%. American-made parts dropped from 11% to 1%.

Howe takes up this round of opportunities. Nomura Oriental International Securities has analyzed that in 2020, Weir shares (Howe) 32 million and 48 million pixel products will be purchased in large quantities by Huawei and OPPO. In the same year, the Xiaomi Mi 10 Extreme Commemorative Edition high-end flagship machine was launched. The rear main camera was equipped with Howe OV48C products. Howe began to return from the secondary camera to the main camera position of the flagship, and gradually changed from the spare tire.

There are also Gekewei and Shengbang shares who have grasped the trend of the times. Gekewei mainly focuses on the domestic low-end CMOS market. In 2010, domestic mobile phones rose, and Gekewei has developed rapidly as a supplier of Xiaomi, OPPO, vivo and Transsion. Gekewei has not been absent from this wave of domestic replacement bonuses. For example, the rear sub-photograph of the Xiaomi Mi 10 Youth Edition has introduced Gekewei GC02M1 products.

In terms of quantity, Gekewei’s shipment volume is very large, even surpassing Sony, and it can sell more than 100 million pieces in a month. However, because the unit price is too low, it is sometimes classified in the “Others” column on the statistical graph of the market calculated by the sales amount. It can be seen that in the field of image sensors, the high unit price brought about by advanced technology is an overwhelming advantage.

But Gekewei also tried to make changes. In July 2020, Gekewei applied for a listing on the Science and Technology Innovation Board to raise 6.96 billion yuan. This money mainly did two things: building a 12-inch integrated circuit project to increase production capacity and developing higher-end CMOS products. It is hard to say whether Gekewei can switch from the low-end to the high-end, but it is ready to undertake this wave of domestic substitution opportunities.

There is also Shengbang Co., Ltd., which is a power management chip. Its LED drivers, operational amplifiers, unlocking motor drives and other products are used in mobile phone electronics, automobiles and medical fields. The products are not only independently researched and developed, but their performance is comparable to foreign counterparts and domestic substitutes. Very big.

It is precisely because Huawei triggered this wave of domestic localization that many domestic companies have been able to enter the high-end market. It is basically difficult for companies to enter the high-end market by relying on their own technology research and development to catch up with foreign manufacturers.

Blacksmith need its own hardware

Most of the development paths of domestic electronics companies start from the low-end and then transform to the high-end. It’s just that there are not many successful cases of this kind of growth path that the company wants to come up with.

To achieve a low to high, it is a relatively feasible choice to rely on the opportunities of the times given by domestic high-end enterprises. Howe, Gekewei, and Shengbang are all examples. But we must clearly realize that when opportunities come, do companies have the core ability to seize them? Ironing still needs to be the key to all development.

As more and more Chinese brands are born, the electronics industry will develop stronger, and long-term investment in basic skills will enable Chinese companies to better embark on the path of shifting from the division of labor to the value chain.

Reference materials:

[1] A single chapter of the core character, Love Jiwei

[2] Weir shares: the leader in image sensor chips, Nomura Orient International Securities

[3] Weir shares: Taking advantage of domestically produced to replace Dongfeng, CIS leader continues to grow, China Galaxy Securities

[4] The Nobel Prize Winner in Physics: The Hard Road from Theory to Practice, Science and Technology Daily

[5] Weir shares: triple power boost, accelerated rise of CIS leader, Dongxing Securities

This article comes from WeChat public account : Yuanchuan Technology Review (ID: kechuangych) , author: Li Jianhua

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-07-26 12:48
Next 2021-07-27 03:09

Related articles