Qualcomm's Snapdragon Wear Elite chip is made for smartwatches and AI devices
Qualcomm's Snapdragon Elite chips are reserved for the best Android phones and laptops, and now the company has introduced the first in the Elite series for wearables. The Snapdragon Wear Elite processor is designed for smartwatches and AI devices like pendants and promises up to a fivefold increase in single-thread CPU performance, Qualcomm announced.
The new processor is built on a 3nm process to improve speed and efficiency over previous models, while boosting the number of cores to five (one big core at 2.1GHz and 4 little cores at 1.9GHz). With those changes, the company is promising up to five times faster single-threaded performance, with GPU speeds boosted up to seven times.
Qualcomm
The Snapdragon Wear Elite is also equipped with a new NPU that allows low-power AI use cases like keyword recognition along with noise cancellation. It's also the first Snapdragon wearable processor with a dedicated Hexagon NPU supporting AI models with two billion parameters. That will allow new "personal AI experiences," the company said, like context-aware recommendations, natural voice interactions, life logging and AI agents that can orchestrate tasks on your behalf.
Wear OS devices with the chip should see up to 30 percent improved battery life and charging speeds of up to 50 percent in ten minutes. It also allows for more types of connectivity, including 5G reduced capability, micro-power Wi-Fi, NB-NTN for satellites, Bluetooth 6.0, GNSS and UWB. However, manufacturers will be able to source versions of the chip without some of those wireless features.
Whether the Snapdragon Wear Elite will give Wear OS watch manufacturers a better chance to chip into the 50-plus percent market share of Apple's Watch remains to be seen. The first devices using the chip will start to ship in the "next few months," Qualcomm said. "Leading global partners are supporting the platform including Google, Motorola and Samsung."
This article originally appeared on Engadget at https://www.eng
Clustered
