20 years of Chinese TV chips
Latest update time：2019-08-18
Reads：
Honor Smart Screen "is not a TV, but the future of TV."
When consumers heard or saw the "tricky" hype, they were all a little surprised, and then they kept Honor's smart screen in mind.
This is not the first attempt at smart TV.
When Lei Jun acquired Duokan at the end of 2012, developed the Xiaomi Box, and then launched the Xiaomi TV, the traditional TV industry had already felt the "wolf coming".
At that time, traditional TV manufacturers only used "intelligence" as a gimmick to increase brand value, but the TV was not actually "intelligent".
Wang Chuan, who admired Apple's products, designed the remote control to have "only 11 buttons" and borrowed the software system of Apple TV to make Xiaomi TV a hit.
For a long time, Huawei's army fought in another battlefield, but it was connected to the TV battlefield.
HiSilicon's set-top box chips started in 2008, and in 2011, it gained product advantages over foreign manufacturers such as ST (STMicroelectronics) and Broadcom in the telecom operator field. Now it has become the largest player in the operator IPTV field.
From set-top box chips to TV chips, combined with its own operating system, Huawei has unified the industry chain from top to bottom, from the basic layer to the application layer.
Such top-down innovation does not make the smart screen look more "intelligent" than other smart TVs, but the underlying innovation initiated by the chip layer is causing Huawei to increasingly erode the "Internet dividend" that originally belonged to other manufacturers.
As a market that has been developing for 20 years, set-top boxes and TVs have always been a battleground for chip makers.
Domestic chips started in the low-end set-top box market driven by domestic policies. After proving their capabilities, they gradually penetrated into the high-end set-top box market and were able to go abroad to compete with international giants.
In the era of smart TVs, domestic chip manufacturers have taken a step further and, taking advantage of the structural changes in the TV chip market, have entered the more difficult TV chip market.
From a regulated market to a free market, how can domestic chip manufacturers make steady progress, achieve domestic substitution, and then go global to lead the transformation of market segments?
TV Chip Wars provides us with the best example of the rise of Chinese chips.
Alchemy of "underground chips"
In 1999, on the occasion of the 50th anniversary of the founding of the People's Republic of China, in order to meet the demand for high-definition video broadcasting, the government mobilized a number of scientific research institutes including Zhejiang University to participate in research and development, thus laying the technical and talent foundation for China's set-top box chips.
Since then, China has entered a golden period of large-scale popularization of digital television.
Cable TV networks began to be rolled out in cities on a large scale, but because of the high cost and paid programs, Gehua's set-top boxes were given away for free. The company actually collected "landing fees" from TV stations. In
2008, before the opening of the Olympic Games, China launched ChinaSat 9, which adopted the "clear stream without encryption" method to enable rural areas to directly receive programs. This is the "Village-to-Village" project.
The third generation of the “Village-to-Village” project—“House-to-House”
At that time, the Radio and Television Administration designated Hangzhou Guoxin and Hunan Guokewei to develop and mass-produce set-top box chips, and stipulated that chip manufacturers could only sell them to designated manufacturers.
However, some chip companies violated the regulations and sold such chips to non-designated manufacturers, resulting in a large number of "copycat boxes" using "black chips" on the market.
ChinaSat 9 is not encrypted. As long as there is a chip, the rest of the manufacturing process is not a problem.
Counterfeit boxes are very popular. In 2010, 80% of China's 80 million live broadcast satellite users used counterfeit products.
To deal with this problem, the first generation of Village-to-Village Communication has been upgraded many times without success, and the only way is to reduce the number of stations to gradually reduce the appeal of counterfeit boxes.
A group of IC designers that make set-top box chips made their first fortune in this wave of "black chip" craze.
Wang Kai is a veteran in the chip industry, and he still remembers the situation at that time.
He recalled to CV Intelligence that after the Radio and Television Bureau issued a ban on some chip manufacturers in 2010, the production and circulation of "black chips" had not actually stopped.
Some companies set up shell companies to hide their profits through the form of "distributors."
However, over-reliance on the underground market has had a negative impact on the continued growth of some chip manufacturers.
After learning from the lessons of the first generation of ChinaSat 9, the Radio and Television Administration introduced the CA giant NDS to encrypt satellite programs.
NDS's encryption system rarely undergoes large-scale cracking. After being banned from the market by the Radio and Television Administration,
some chip manufacturers continued to rely on the sales performance of "black chips" and did not continue to invest in research and development, thus losing growth opportunities
.
Set-top box products with CA are a steadily growing market both at home and abroad.
Wang Kai believes that
there are big differences in the functionality of televisions in the domestic and foreign markets
.
Domestic televisions are more for cultural propaganda, so people are reluctant to spend money on set-top boxes at first.
But in foreign countries, TV bills are listed separately in consumers' monthly expenses because consumers regard buying TV programs as a pure entertainment activity.
In order to ensure that the paid programs are not cracked, set-top boxes with CA are a rigid demand product.
While its peers were obsessed with selling "black chips" to make money, Hangzhou Guoxin made great efforts in this field and sold its chips to overseas markets.
This was also a wise choice:
"Black chips" represent the needs of low-end and mid-range users, and this part of the demand is shrinking.
In sharp contrast to low-end set-top boxes, with the rapid development of China's economy and the rapid growth of people's average income, pay TV is becoming more and more popular, and telecom operators are also accelerating their entry into the set-top box market to get a share of the huge TV market.
Huawei's opportunity has come.
IPTV's Great Leap Forward
While radio and television are developing their own television business in full swing, IPTV is also gaining momentum, and China Telecom (formerly known as Netcom) and radio and television are both preparing to make great strides in this field.
IPTV, or interactive network television, has very high requirements for the performance of the transmission network.
Unlike traditional set-top boxes, which only need to meet the availability but are highly sensitive to cost, IPTV has very high requirements for performance, and operators are not so sensitive to cost, so this segment is more profitable
.
According to an article published by former Huawei employee Dai Hui, after HiSilicon was founded in 2004, the most important thing that the current Huawei Fellow Ai Wei did with the head of the software company's digital entertainment product line and the head of marketing was "product definition." They
brainstormed for a month or two before they finally decided on this matter.
During the product definition process, the focus of the debate was whether to support H.264 HD.
Supporting H.264 HD was ahead of most standard-definition products at the time,
but the person in charge of the product line insisted on making this request.
Because it would take two or three years to successfully tape out and launch it on a large scale, considering the product life cycle of at least two or three years.
Defining the specifications of a chip requires at least predicting the terminal product trends in the next five years in advance.
Wang Kai believes that the core competitiveness of HiSilicon in developing set-top box chips is nothing else but "
product definition
".
Before making a chip, set-top box manufacturers are not familiar with the terminal market demand. They just follow the routine and use whatever chips ST and Broadcom give them.
If you ask operators what kind of chips to make, they don't know much.
Copying the existing product specifications of ST and Broadcom is labor-saving, but by the time the chips are out, such products may have been eliminated.
Therefore, the "product definition" in chip design can almost be said to be a somewhat "mysterious" job.
The work of "product definition" lies at the intersection of science and art, and companies that have mastered this skill always seem to be able to set the trend for a while
.
The iMac, iPod, iPhone and Apple Watch launched by Apple successively were not groundbreaking products, but relying on Jobs' genius intuition, the products defined by Apple have always been able to lead the global consumer electronics trends as a "latecomer innovation".
Of course, it is not enough to just define a product after making a chip.
After "product definition", the technical requirements are transformed. In the path of technical implementation, the architect plays the most important role. He needs to coordinate multiple key technical factors such as processors, storage requirements, bus structure, etc. according to the needs of product managers, and make trade-offs between development time, flexibility and cost.
After the architecture is built, the next step is to "build IP", that is, SoC - the process of integrating various functions.
SoC integration is not an easy task, and the fact that there are only a handful of IC designers of scale in the audio and video field in China is a proof of this.
When starting to make set-top box SoC main control chips, most domestic IC designers generally outsource most IPs to achieve specific functions.
With the accumulation of experience and improvement of architecture, IC designers will gradually consider making their own IPs to achieve functional integration, reduce chip area and reduce R&D costs.
After setting up the architecture and assembling the IP, the work on the set-top box chip is not over yet.
To complement the chip hardware team, there is also a software team of considerable size.
Wang Kai told CV Intelligence that for companies like Rockchip that make master AP chips,
the ratio of software to hardware team personnel is 1.5-2:
1,
which means that if the company has 100 people working on chip hardware, it needs to have 150 to 200 people working on software.
The software team of a chip company is far different from that of an Internet company.
Existing mobile Internet companies all use upper-level languages such as Java, which have no impact on the underlying hardware such as chips.
However, chip companies write software using low-level languages, which require consideration of factors such as the chip's computing efficiency and energy consumption. The learning process is long, and it takes two to three years to become a master, and five years to become proficient. However, after working for ten years, the salary of an Internet software engineer of three to five years is too different due to the huge difference in demand.
At the time when HiSilicon was making chips, domestic talent was very scarce, not to mention the high-performance chips that HiSilicon wanted to make.
HiSilicon poached a group of experienced Chinese engineers from Broadcom, laying the foundation for basic talent reserves.
With a strong talent pool, accurate market foresight, and successful tape-out, HiSilicon chips received an urgent order for 100,000 set-top boxes from Guangdong Telecom in 2008.
At first, Guangdong Telecom's first choice was Broadcom, but due to Broadcom's long chip supply cycle and Guangdong Telecom's unpredictable IPTV promotion progress, HiSilicon chips got the opportunity to debut.
With HiSilicon chips not inferior to Broadcom in performance tests, HiSilicon obtained access to this market.
According to Wang Kai's recollection, HiSilicon first achieved market shipments of tens of millions in the standard-definition DTV market of radio and television, starting with low-end set-top boxes.
In this discrete market divided by regions, Broadcom used the MIPS CPU architecture and was slow to move to the ARM architecture. The software ecosystem was not perfect, and set-top box manufacturers had too much work to do, so the product was not popular.
Subsequently, in the operator relationship market where Huawei has always been relatively strong, HiSilicon fully leveraged its advantage in advanced "product definition" to achieve a faster pace and higher performance than foreign chip manufacturers, thereby winning a large order for telecom IPTV.
The growth path of HiSilicon's set-top box chips is the general path of domestic chip manufacturers.
Domestic chip manufacturers start from the low-end market, prove their reliability and high performance in a semi-policy and semi-free market, and grow based on the cultivation of the domestic market. Then they have the opportunity to climb up step by step and then launch an attack on the global market.
Smart TV melee
In 2012, the “smart” trend swept the television industry, and Xiaomi and LeTV took advantage of the opportunity.
BOE has made its own low-cost LCD screens in China, and the Japanese screens that Sony and Sharp were proud of no longer have an advantage;
the Internet is becoming the birthplace of self-made entertainment content, and people are increasingly interested in paying to watch Internet content.
In short, the TV industry chain is being rapidly reshaped.
Internet companies have seen business opportunities in the “transformation” of television.
Before Internet TV, TV was more of a bulky household appliance, bulky, difficult to use, and with a decreasing power-on rate. After Internet TV
, TV has gained a new lease of life, with an Android operating experience, a massive amount of Internet content viewing experience, and a user-friendly UI interface. TV is changing from a bulky household appliance with a life cycle of five to eight years to an electronic fast-moving consumer product with a life cycle of only two to three years.
"Buy a membership and get a TV for free" and "Buy a TV at a super low price to watch the interstitial ads" are both ways in which Internet companies "sell" hardware at a low price in order to seek software profits.
The operation is as fierce as a tiger, and the new punch kills the old master.
Xiaomi TV cannot escape the usual "Xiaomi-style" criticism. The underlying system is Android, the chip is purchased from outside, and the gameplay of MIUI TV is similar to "the Chinese version of Apple TV".
But even so, it does not prevent consumers from chasing after Xiaomi TV:
According to data from iiMedia Research, in the first half of this year, Xiaomi has become the best-selling TV brand in China, shipping about 4.2 million units, accounting for 20% of the market, leading the second-place Hisense by 900,000 units.
The rapidly iterating hardware products and software systems, as well as the mushrooming of various Internet TVs, have forced chip manufacturers to produce rapidly iterating, high-quality, low-cost, and widely applicable chip products.
OTT Internet boxes that are lightweight and have low user conversion costs were the first to explode. Shanghai IC design company Amlogic, Rockchip and Allwinner, which originally made tablet main control AP chips, have all flocked to this market.
With its capital and technology accumulation in the field of hard disk players and systematic optimization of Internet TV, Amlogic's products sell much better than Rockchip and Allwinner's. Rockchip and Allwinner
's chips are just the main control chips of tablets with more complex functions cut out and used without systematic optimization and close follow-up improvement.
In the field of Internet TV chips, M-star, a long-established Taiwanese IC design company, maintains its consistent advantages.
Unlike set-top box chips that mainly handle audio and video encoding and decoding and operating systems, TV chips process the audio and video content sent by the set-top box. Because the audio and video content contains multiple types of analog signals, it requires more experienced engineers and takes longer to accumulate experience.
Since the beginning of the century, Morningstar has been fighting with TV chip manufacturers in Europe, the United States, and Japan, and has taken a dominant position in the field of TV chips.
Morningstar has been working on TV chips for a long time, and all IPs are self-made, so it is good at integration, improvement and cost reduction. The single-chip integration capabilities of TV chips and main control chips have helped Internet TVs to significantly reduce costs.
Xiaomi TV initially used two chips, one was a Qualcomm processor and the other was a Morningstar TV chip. However, due to the high cost of Qualcomm chips and their inferior performance to Morningstar's integrated single chip, the second and third generations adopted Morningstar's chip solution.
The merger of MediaTek and MStar, approved by China's Ministry of Commerce in 2013, provided an opportunity for domestic chip manufacturers to enter the field of TV chips.
Before that, some domestic chip manufacturers were already making TV chips, but they were completely blocked by the main and secondary chip solutions of MStar and MediaTek.
After Morningstar was merged into a company by MediaTek, TV manufacturers had to find another chip manufacturer as a secondary solution provider.
Based on the previous experience in Android software development, HiSilicon's smart TV chip won a time difference advantage, thus obtaining a valuable entry ticket.
Chip manufacturers from mainland China are competing with Taiwanese chip manufacturers, while foreign chip manufacturers are attacked from all sides and are losing ground.
In 2013, ST still had a 30% share of the Chinese set-top box market, and the head of the China region was confident that it would maintain this market share.
In early 2016, when the market share was still very impressive, ST decisively withdrew from the market:
1,400 business unit employees were laid off and 600 employees were transferred.
The huge global market created by ST's withdrawal was quickly filled by Morningstar.
While Morningstar was chasing the global high-end market, IC designers in mainland China quickly filled the mid- and low-end market.
Wang Kai found it "unbelievable" that the company closed down its stores because it was not profitable enough when its market share was very high. Perhaps only Europeans could do such a thing. For any Chinese manufacturer, such a business decision would seem unreasonable.
The smart TV has created new business opportunities. In the midst of industry changes, domestic chip manufacturers have quickly occupied most of the OTT box market and entered the field of TV chips.
However, the business trend dominated by Internet companies has invisibly "alienated" the functional experience of TV, and hardware providers have fallen into decline.
Who can reverse the alienated soft and hard structure of Internet TV?
From the core to the bottom
At present, Huawei is trying to migrate the complementary relationship between Kirin chips and mobile phones to TVs, and the Honor Smart Screen is becoming a training ground for HiSilicon's self-developed TV chips.
However, at the launch of Honor Smart Screen, the most eye-catching thing was Hongmeng OS, which also attracted the most criticism.
On the contrary, the three self-developed chips such as Honghu did not stir up any waves or arouse any doubts.
So, the question is, where is the premium of this "screen" that claims not to be a TV?
Compared with Xiaomi, whose prices are often very attractive, the premium of Huawei's products is obvious.
Xiaomi is a believer in Apple, but the profit model created by this company is almost entirely Internet-based, which is in sharp contrast to Apple's reliance on high hardware and operating system premiums to make a big profit.
Now, Huawei seems to be taking an Apple-like route, whether actively or passively. Its
self-developed TV chips and Hongmeng OS have become a genuine Chinese version of Apple TV, rather than a "Chinese version" of Apple TV.
As a latecomer, Honor Smart Screen may not be able to create too many novel experiences, after all, Internet TV has long existed.
Huawei's move is more like integrating the strength of its self-developed chips and operating systems with the user base of smartphones to obtain the certain dividends of the mature Internet TV market.
Entering the smart TV market will inevitably affect HiSilicon's exported TV chips, but given that HiSilicon's share of smart TV chips is not large, only Huawei's own terminals can use the chips.
Xiaomi started from the bottom up with MIUI, and Huawei started from the top down with three self-developed chips.
The battle between the two major routes of Internet TV, "Internet dividend" and "hardware dividend", is bound to be unavoidable.
*Disclaimer: This article is originally written by the author. The content of the article is the author's personal opinion. Semiconductor Industry Observer reprints it only to convey a different point of view. It does not mean that Semiconductor Industry Observer agrees or supports this point of view. If you have any objections, please contact Semiconductor Industry Observer.
Today is the 2040th issue of content shared by "Semiconductor Industry Observer" for you, welcome to follow.
Recommended Reading
★ New trends of Alibaba's Pingtouge chip
★ Japan and South Korea break up, who will be most hurt in the semiconductor industry?
★ Unveiling the Secret of Tianji Core
Semiconductor Industry Observation
" The first vertical media in semiconductor industry "
Real-time professional original depth
Scan the QR code , reply to the keywords below, and read more
RF｜TSMC｜Jiangbei New District｜AMD｜FPGA｜EDA｜IC｜MRAM
Reply
Submit your article
and read "How to become a member of "Semiconductor Industry Observer""
Reply Search and you can easily find other articles that interest you!
京公网安备 11010802033920号