Recently, foreign media Winfuture revealed Qualcomm's new Wear 5100 series of smart wearable chips, including the Wear 5100 and Wear 5100+. The difference lies in the latter's integration of the QCC5100 coprocessor, used to improve the smartwatch's battery life in low-power mode. Regarding the manufacturing process, a key concern for users, this series of chips will skip multiple process nodes, jumping directly from the previous generation's 12nm to 4nm process, keeping pace with the development of flagship mobile phone SoCs.
Compared to its predecessor, the Qualcomm Wear 5100 series represents a significant leap forward in manufacturing process technology. For many manufacturers looking to improve the battery life of their smartwatches, the Wear 5100 series may be a promising new option.
Achieving both high performance and low power consumption?
In the smartwatch industry, there are roughly two different product development directions: one focuses on long battery life, while the other emphasizes the "fully intelligent" experience of smartwatches. A major factor contributing to this situation is the chip. Even if some manufacturers want to pursue both performance and battery life, they are limited by the computing power and power consumption of chips, leaving them with no choice but to either choose low-power chips to make lightweight smartwatches or choose Qualcomm's Wear series chips to develop fully intelligent smartwatches.
Last September, rumors circulated that the Wear 5100 series would use an A73+A53 architecture, focusing on improving the chip's performance ceiling. However, according to the latest information, Qualcomm seems to have adjusted its development strategy for the Wear series chips, intending to improve the chip's power consumption.
The Qualcomm Wear 5100 series features four A53 cores (up to 1.7GHz) and an Adreno 702 GPU. It supports eMMC 5.1 flash storage and 4GB of LPDDR4X RAM. Compared to the previous generation Wear 4100 series, the CPU core specifications remain the same, but the Wear 5100 series has received significant upgrades in GPU and flash storage. Samsung's 4nm manufacturing process is undoubtedly the core of this upgrade. On paper, it's clear that Qualcomm didn't blindly pile on specifications, but rather aimed to improve chip performance and reduce power consumption by changing the manufacturing process, attempting to address the issue of weak battery life in smartwatches.
In addition, Xiao Lei also discovered some new features of the Wear 5100 series. Regarding camera support, the Wear 5100 series supports dual-camera setups, with sensors supporting up to 13MP and 16MP respectively. Furthermore, when using a single camera, it supports recording 1080P video. Qualcomm may be aiming to promote the diversification of smartwatches.
Similar to smartphones, many new features and functionalities of smartwatches require chip support. Take the imaging capabilities of smartwatches as an example: smartwatches developed for students often feature front-facing cameras. The arrival of the Wear 5100 series not only improved the video recording quality of smartwatches but also allowed some manufacturers to expand the possibilities of smartwatches by using a combination of a front-facing main camera and an ultra-wide-angle lens.
However, it's worth noting that even the Wear 5100 series, which uses a 4nm manufacturing process, struggles to fundamentally improve the battery life of smartwatches. The Samsung Galaxy Watch4, which already uses a 5nm chip, can only last two days on a single charge under normal use.
The battery life of a smartwatch is closely related to system scheduling. Manufacturers using fully intelligent operating systems like Android and Wear OS, who want to improve both performance and battery life, currently have to incorporate a complete low-power chip. In my opinion, the Wear 5100 series, using a 4nm process, offers smartwatch manufacturers a new option, allowing them to move beyond the more power-intensive 12nm or even 28nm chips and further optimize for improved battery performance.
The smart wearable chip industry also needs internal competition.
In the past few years, apart from Apple and Samsung, other manufacturers did not fully commit to the smart wearable chip field. This is reflected in the application of manufacturing processes: Qualcomm's Wear 4100 series released in 2020 and Wear 3100 released in 2018 used 12nm and 28nm processes respectively, while Unisoc's W307 and Rockchip's RK2108D also used 28nm processes.
The main reason why manufacturers have not adopted advanced manufacturing processes for smart wearable chips in a timely manner is that upstream smart wearable chip suppliers are unwilling to take risks. Without clear and sufficient market demand, they are reluctant to directly choose more expensive advanced manufacturing processes. In the early and mid-stages of smartwatch development, mobile phone manufacturers themselves did not clearly define the product positioning of smartwatches (e.g., launching some immature products) and did not fully understand the actual needs of users. This led upstream smart wearable chip suppliers to choose compromise solutions or maintain a slower product update pace.
Taking the OPPO Watch2 series as an example, since the upstream supply chain chips are slow to iterate and their performance and power consumption are difficult to meet the design requirements of the new products, in order to balance performance and battery life, we can only take a different approach and use the "1+1" dual-chip solution. Qualcomm's Wear 4100 and the low-power Apollo 4s chip are used on one smartwatch, allowing users to switch the chip usage solution according to different usage scenarios.
Apple and Samsung's advantage lies in their dual role as players in the smartwatch market and their ability to develop their own chips. This gives them a stronger market insight and resilience compared to their upstream supply chains, allowing them to launch 7nm or 5nm wearable chips more quickly. In other words, their integrated "production, education, research, and sales" model enables faster market response, and their product updates don't depend on the whims of upstream chip manufacturers.
The intense competition within the smart wearable chip industry is helping to drive the positive development of the entire smartwatch industry. In the Apple Watch Series 6 released in 2020, Apple developed a dedicated S6 chip based on two small cores of the A13 (7nm). Samsung, not to be outdone, released the Exynos W920 last year, which was manufactured using its own 5nm process.
In contrast, Qualcomm's Wear 4100 series, still using a 12nm process, is several generations behind in manufacturing technology, resulting in inferior chip performance and power consumption. Manufacturers that only update their products after Qualcomm releases new chips are often forced to temporarily halt product updates or release "light smartwatches" emphasizing long battery life. To regain the confidence of its partners, Qualcomm may have no choice but to skip multiple process nodes and use a 4nm process for its new chips.
Based on data from Apple and Samsung's released chips, as well as the leaked information about Qualcomm's Wear 5100 series chips, it's clear that flagship smart wearable chips will keep pace with the development of mobile SoCs in terms of manufacturing processes. In my opinion, whether it's smart wearable chips or the computing units of TWS earphones, the use of advanced manufacturing processes is an inevitable trend. Changing product usage scenarios and user needs are driving manufacturers to adopt newer processes. After all, 60% of chip performance improvements come from advancements in manufacturing processes, and 40% from design. For these "small chips," using new manufacturing processes is the fastest way to improve overall chip capabilities.
Is developing your own chips the ultimate goal?
Last year, major domestic mobile phone manufacturers accelerated the development of their self-developed chips, releasing ISPs or NPUs for mobile phone photography. In the field of smart wearables, Huami released the Huangshan 2S chip based on the RISC-V architecture. In fact, in the technology industry, manufacturers who want to improve the core competitiveness and premium of their products often have no choice but to embark on the path of self-developed chips.
Compared to mobile SoCs, the development of smart wearable chips is less difficult, as they do not require the latest ARM CPU and GPU architectures. As for the baseband, which troubles many manufacturers, eSIM-enabled smartwatches are not a strong demand. For many users, it's not a common practice to go out with only a smartwatch and no phone. Manufacturers could simply develop Bluetooth versions first to capture a segment of the user base.
Secondly, smart wearable chips are quite similar to mobile phone SoCs, allowing domestic manufacturers to accumulate experience and patents in the design and development of smart wearable chips, thus driving the construction of their self-developed chip systems from the bottom up. For manufacturers who want to cultivate the smartwatch field, in the long run, only by launching their own smart wearable chips can they have the opportunity to surpass Apple and Samsung in market share and profit margins, and control the product iteration and core selling points.
The chip iterations in smartphones and watches over the past two years reveal that relying solely on upstream supply chains for chips is too passive. Either the chips are not updated in time, or they have various problems, making it impossible to support the high-end development of the products.
Of course, compared to the relatively simple ISP, the internal structure of smart wearable chips is more complex, involving different chip design fields. It even requires SiP packaging technology to improve chip integration, an area where many mainstream domestic manufacturers lack sufficient experience. Apple and Samsung's ability to launch high-performance smart wearable chips is mainly due to their technological accumulation in the mobile SoC field. I also believe that many domestic manufacturers in the technology sector will eventually embark on the path of self-developed chips. With increasing market competition, most of the manufacturers that survive will possess certain chip design capabilities.
