April 19, 2021

Rockchip launches smart wearable chip RK2108D “dual standby” ultra-low power consumption design

Recently, Rockchip officially launched a new generation of high-performance and low-power smart wearable chip RK2108D, which adopts 28nm process design, dual-core architecture, and features high clock speed, large memory, and low power consumption. In response to the product requirements of smart wearable products in terms of power consumption, response speed, voice recognition and operating system adaptation, Rockchip’s RK2108D solution has achieved significant and effective technical optimization.

  1. Unique “dual standby” low power consumption mode, shortening the time of lifting the wrist and lighting the screen by about 30%

 

Rockchip RK2108D supports self-developed primary standby + deep standby “dual standby” mode, and its power consumption is industry-leading. At the same time, through actual measurement, the power consumption of the chip is 0 in the deep standby mode. From deep standby to wake up and bright screen, other solutions usually take about 300-600ms. RK2108D relies on M4F with a frequency of up to 400MHz, which only takes about 200ms, and the time is shortened by about 30%, providing a smoother and more convenient user experience.

  1. The response speed is greatly improved, and the touch feel is more sensitive

 

RK2108D adopts HiFi3 DSP + ARM M4F dual-core architecture, based on high-performance DSP, RK2108D has 2D rendering operator acceleration capability; after actual measurement, when the second hand of the watch rotates by 1 division, other Arm rendering schemes usually take about 10ms, and the rendering time of RK2108D is only about 1ms. Significantly improve the response speed of various functions of the product.

RK2108D has a high-speed MIPI display interface, supports hardware layer overlay, and supports a maximum refreshing frame rate of 60FPS per second. Most other solutions only support 30FPS; when sliding the dial or performing other manipulation actions, products based on the RK2108D solution have a better touch feel Sensitive, not stuck.

 

  1. With its own AI speech recognition engine, the cost of man-machine dialogue is low

 

RK2108D built-in HiFi3 DSP with 600MHz frequency, providing high-quality voice signal processing capabilities. With its own AI speech recognition engine, it can realize the AI ​​speech recognition function without additional hardware, which is more cost-effective.

  1. The development is more stable and supports rapid UI transplantation

 

RK2108D supports the adaptation of RT-Thread domestic operating system, is mature and stable, and has a good software ecology, which is convenient for developers to develop various applications. At the same time, RK2108D uses the LittlevGL UI framework to support developers to quickly transplant existing UI to new products, effectively shortening the product development and design cycle.