June 13, 2024
news
During COMPUTEX, Fibocom launched a series of on-device AI solutions powered by Qualcomm QCS8550 and QCM6490 processors. The platform is developed to support the development of compute-intensive applications such as robotics, autonomous vehicles, video collaboration and smart retail, and to facilitate the empowerment of industrial digitalization and intelligent transformation.
Fibocom's flagship on-device AI solution delivers enhanced performance and unparalleled multimedia capabilities. Equipped with an octa-core CPU and Adreno 740 GPU, it supports up to four simultaneous displays and 8K video encoding and decoding.
It is ideal for use cases that require high-resolution video playback, high-speed data analytics, and low latency, including autonomous vehicles, robotics, remote medical surgery, computer vision systems, live streaming, and video conferencing systems.
The solution features an octa-core processor with high-speed HVX (Hexagon Vector Extension) technology and a high-performance graphics engine for 4K video playback and multi-channel camera input. It can run up to 5 ISPs (Image Signal Processing) and up to 5-8 camera streams simultaneously, which is useful for multi-camera deployments and dual screen display scenarios.
Flexible wireless connectivity includes 5G, Wi-Fi and Bluetooth, and it has a built-in GNSS receiver for precise navigation indoors and outdoors.
“Mastering the productivity of AI and creating value-added solutions from the edge side is crucial for our customers who are in urgent need of building smart devices based on our solutions. We are pleased to develop these solutions based on Qualcomm Technologies' advanced and powerful chipsets, which not only provide the basic architecture for edge intelligence but also enhance the flexibility of network connections such as 5G/Wi-Fi/Bluetooth,” said Ralph Chao, vice president of MC BU at Fibocom.
It supports popular operating systems such as Android, Linux, and Windows. The platform provides up to 13 TOPS of computing power, helping you efficiently handle data-intensive calculations and processing while running a variety of 1.3B/3B/7B open source LLMs on your device, making it ideal for smart retail, in-vehicle infotainment (IVI), and industrial inspection.
“This collaboration is a testament to our commitment to advancing AI capabilities at the edge, improving performance and efficiency for a variety of applications from industrial automation to smart retail,” said Dev Singh, vice president of business development and head of Building, Enterprise and Industrial Automation, Qualcomm Technologies, Inc.
For more information, visit fibocom.com.
Chad Cox is a production editor at Embedded Computing Design, covering the news cycle, newsletters, social media and advertising. He has a BA in Cultural Analysis and Literature from the University of Cincinnati.
More from Chad
