21st Century Business Herald reporter Shi Enze reported from Shenzhen

As the entrance to AI, visual and audio have become a battleground in the consumer electronics industry chain. As the core chip manufacturer in consumer electronics, it is natural to follow the news.

Recently, the 21st Century Business Herald reporter communicated with many domestic audio chip manufacturers at the "International Integrated Circuit Exhibition and Seminar" and found that in the face of the tuyere of AI headsets, they did not rush to pursue "big" as at the beginning of the birth of the domestic AI model, that is, the general scale effect.

On the contrary, they are more inclined to take the path of specialization and verticalization, and the real efforts they have accumulated over the years are actually on products and scenario solutions.

The consideration for this is that, on the one hand, it can cope with the trend of increasingly fierce price competition in domestic chips, and on the other hand, the speed of change of AI is too fast, and too much follow-up will become uncompetitive in the market.

Of course, there is also another faction of domestic manufacturers in the market who firmly follow the "white label" route, that is, the pursuit of scale effect. In the past two years, these manufacturers are actively expanding overseas, especially to the Southeast Asian market. Some manufacturers have also enjoyed the dividends of the times on this road.

But in any case, the war for wearable AI products, represented by headphones, has just begun.

The third way beyond branding and white labels

Eight years ago, in September 2016, the first generation of AirPods was officially released.

For this kind of "completely wireless earphones" without wires, even if Apple does it, it has been criticized a lot once it is launched. However, under all kinds of criticism, AirPods conquered the wireless headphone industry, and even single-handedly changed the pattern of the headphone market.

From 2016 to 2018, TWS earphones became the fastest-growing segment of wearable devices in the world, with a Futuresource report pointing out that in 2018, AirPods accounted for 40% of the global TWS earphone market size ($5.4 billion).

Under the wave of TWS brought by Apple, many giants have also been born in domestic wireless audio chips. Among them, manufacturers also have their own magic methods.

As the first TWS earphone chip, Hengxuan Technology (688608. SH) has been the most notable in recent years. net profit attributable to shareholders of listed companies in the first three quarters of 2024 was 289 million yuan, a year-on-year increase of 145.47%; The non-net profit after deduction exceeded the level of the whole year of 2023 in the middle of the year.

Another in the "white label" TWS Bluetooth headset market is calling for wind and rain (688332.SH) is in the cost-effective, occupy a place. In the first half of 2024, the company's net profit attributable to the parent company in the first three quarters exceeded 200 million yuan, and the annual net profit growth rate is expected to reach more than 20%.

In addition to the above two distinctive manufacturers, the positioning of other domestic chip factories is in between. The Android market retains the blank of TWS earphones, which also gives the market room to accommodate the third way beyond the brand and the white label.

On this third road, domestic players have been deeply cultivated in the vertical field with their own abilities, such as UNISOC Zhanrui, which has the same supporting ecology of mobile phones and Bluetooth headsets as Qualcomm, Torch Core, which has many years of accumulation in Bluetooth audio, and Telink Micro, which is among the top three in the world in Bluetooth low consumption (BLE) chip shipments, are all players that have emerged in recent years.

Not a "volume" model, a "volume" scenario

Different from the beginning of the birth of AI large models, everyone has to make general large models. In the category of AI headsets, which is oriented to C-end users, taking the vertical road is more favored by most manufacturers.

At this exhibition, the audio marketing director of a medium-sized Bluetooth chip factory said that the company "basically does not consider doing the general headphone market" at present, because it cannot give full play to the company's original accumulated technical advantages. With the help of AI algorithms, they will explore scenario-based solutions in the future. For example, in response to customer needs, they have developed a hearing aid product in the form of TWS earphones, and accurately segmented the target customer group to 700 million hearing-impaired people around the world.

It is worth mentioning that the advantages of the above-mentioned subdivided scenarios, under the innovation of AI algorithms, can also be superimposed with functions such as translation and noise reduction, so as to meet the needs of more diversified scenarios. Therefore, only by polishing from the vertical category can the road be widened.

In addition, another trend is that with the addition of more domestic brand manufacturers, brands are more sensitive to the price of Bluetooth chips. This further squeezes the space in the white label market.

This means that if the "AI headset" is only a concept, it will be drowned in countless homogeneous products. "At present, many headphones with the concept of AI on the market are actually just adding a voice recognition command to the chip, which cannot be called AI headsets at all." A representative of a Bluetooth chip factory said.

What's more important behind the AI headset is the platform ecosystem. The audio marketing director of the above-mentioned Bluetooth chip factory believes that the real AI headset is inseparable from the service of the cloud. For example, the compatibility with iFLYTEK, the top class in the translation industry, the smoothness of switching between different devices, and so on.

From this point of view, the "volume" scenario will become the foundation of medium-sized chip manufacturers. Although the emergence of AI can help headphones better span different scenarios, the balance of sound quality, latency, computing power, and low consumption depends on the accumulation of various manufacturers in the scene, which is also the opportunity in the future segment.

IDC pointed out in the analysis report that with the improvement of AI translation and other functions, smart headsets equipped with AI large models are expected to improve the premium ability of products and promote the development of the mid-to-high-end market of Bluetooth headsets. And this may be the root of cracking the homogenization of low prices.