Qualcomm has announced a new line of system-on-chip (SoC) offerings designed specifically for the Internet of Things.
Comprising the Qualcomm Vision Intelligence Platform, the offerings include two new SoCs for camera-driven machine learning. Dubbed QCS605 and QCS603, the SoCs are designed for energy and thermal efficiency, and feature Qualcomm’s image signal processor technology, its AI Engine, and an ARM-based multicore CPU, GPU, and vector processor. Other components of the Qualcomm Vision Intelligence Platform include computer vision and machine learning SDKs, as well as security and connectivity solutions.
In a statement announcing the solutions, Qualcomm emphasized the computational and image processing capabilities of the new platform, and their implications for AI functionality. “The architecture supports various high level operating systems – engineered to allow developers and manufacturers to easily build differentiating features into their solutions such as on-device stitching for VR 360 cameras, autonomous robot navigation and obstacle avoidance, and video summarization for action cameras,” the company said.
The QCS603 and QCS605 are now in sampling, with reference designs for a QCS605-based VR 360 camera from Qualcomm and Altek currently available, and reference designs for a QCS603-based industrial security camera in the pipeline for the second half of the year. Meanwhile, Qualcomm says that KEDACOM and Ricoh THETA are currently developing products based on the platform.