Qualcomm announced the launch of its 'Dragonwing Q-2390' and 'IQ-2390' processors on September 2, targeting the commercial, consumer, and industrial edge AI markets. The focus is on enabling AI processing directly on devices, from kiosks and smart appliances to factory controllers, thereby reducing costs and design complexity.
The Q-2390 is a new addition to the Dragonwing Q2 series. It integrates a central processing unit (CPU), graphics processing unit (GPU), on-device AI, and camera, display, and LTE communication capabilities into a single platform. It can be applied in point-of-sale (POS) systems, kiosks, smart appliances, and home robots.
The IQ-2390 is the first product in the newly created industrial IQ2 series. It combines AI and machine vision with Ethernet support for time-sensitive networking (TSN) and real-time processing capabilities. Key applications include programmable logic controllers (PLCs), human-machine interfaces (HMIs), industrial vision systems, and energy management equipment.
Designed to operate in environments ranging from minus 30 degrees to plus 115 degrees Celsius, the IQ-2390 takes into account conditions with vibrations and shocks. This design addresses the demand for factory automation that requires immediate analysis and control on-site rather than sending data to the cloud for processing.
Last February, Qualcomm consolidated its industrial and embedded Internet of Things (IoT) and networking businesses under the 'Dragonwing' brand. Earlier this year, it expanded its industrial edge AI business by acquiring five companies, including Ojentics, Arduino, and Edge Impulse. This marks a shift in its semiconductor business focus from smartphones to factories and energy infrastructure.
Both products feature a quad-core Kryo CPU, Adreno 704 GPU, and a real-time RISC-V microcontroller. They support Linux, Android, and Zephyr operating systems, allowing device manufacturers to choose the software environment based on product applications.
Qualcomm is broadening the scope of edge AI applications by adding both high-performance industrial AI chips and cost-effective product lines. The strategy aims to spread on-device AI to small and medium-sized industrial and commercial equipment by integrating AI computation and connectivity into a single chip.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.