The new 10nm chipsets, QCS605 and QCS603, can be used for devices such as drones, robotic vacuum cleaners, 360-cameras, and more. Both contain Qualcomm’s latest development in camera and artificial intelligence.
Driving the capabilities of these chipsets forward is the Vision Intelligence Platform. The platform uses edge computing to make decisions and take actions near the source of the data — for better security, faster processing, and local control.
Jim Merrick, Marketing Director for IoT at Qualcomm, says:
“The processing of that information could be done in the cloud, but that takes resources and time. Because the camera itself has the intelligence, it can decide how to respond based on what it knows instead of waiting for video data to be sent to the cloud and analysed.
Integrating this technology will also push the IoT ecosystem forward, as developers move away from the cloud and focus on the capabilities of the device.”
Qualcomm’s AI Engine is compatible with popular deep learning frameworks including TensorFlow, Caffe, the Open Neural Network Exchange interchange format, Android Neural Networks API, and the Qualcomm Hexagon Neural Network library. Developers have plenty of options when it comes to their preferences.
All of this requires significant horsepower and the Vision Intelligence Platform aims to rise to the challenge using Qualcomm’s heterogeneous computing. The hardware consists of a powerful 8-core, 64-bit Kryo 300 CPU, Adreno 615 GPU, Hexagon 685 Vector Processor, Spectra 270 ISP, and DSPs for sensors and audio.
Seshu Madhavapeddy, VP of IoT Product Management at Qualcomm, spoke to Engadget and said the company’s work on smartphone cameras helped to improve the performance in IoT devices. Security cameras are the most obvious beneficiary here — with devices able to reduce noise in low light, or even complete darkness.
Products with the new SoCs are expected to launch in the second half of the year.
What are your thoughts on Qualcomm’s new platform and IoT chipsets? Let us know in the comments.
thanks you RSS link