Kevin254
Member
Huami Technologies will hold an AI innovation conference on June 15, and yesterday the company hinted on Sina Weibo that it will launch the next generation of AI chips for wearable devices.
Huami Technology released the world's first AI chip for smart wearable devices-"Huangshan No.1" on September 17, 2018. This is a wearable processor with integrated AI neural network, with four core AI engines: heart biometric engine, ECG, ECGPro, abnormal heart rhythm monitoring engine. Huangshan No. 1 has been used in many watches and straps under Huami's Amazfit brand.
Due to the limitations of the usage scenario, it is more suitable than the X86 architecture of the PC and the ARM architecture of the mobile device. The chip architecture of wearable devices must have ultra-high computing efficiency and ultra-low power consumption.
High performance and low power consumption are two major concerns of smart devices. For smart wearable devices that moniters health, performance will affect the speed of disease recognition, and every 1% increase in the speed of disease recognition can save more lives. But how to effectively improve theperformance while keeping the battery power under low consumption is still a problem.
Huami Technology released the world's first AI chip for smart wearable devices-"Huangshan No.1" on September 17, 2018. This is a wearable processor with integrated AI neural network, with four core AI engines: heart biometric engine, ECG, ECGPro, abnormal heart rhythm monitoring engine. Huangshan No. 1 has been used in many watches and straps under Huami's Amazfit brand.
Due to the limitations of the usage scenario, it is more suitable than the X86 architecture of the PC and the ARM architecture of the mobile device. The chip architecture of wearable devices must have ultra-high computing efficiency and ultra-low power consumption.
High performance and low power consumption are two major concerns of smart devices. For smart wearable devices that moniters health, performance will affect the speed of disease recognition, and every 1% increase in the speed of disease recognition can save more lives. But how to effectively improve theperformance while keeping the battery power under low consumption is still a problem.
