Firefly has launched a series of intelligent products based on the Qualcomm QCS8550 chip, including system on modules, motherboards, edge computing hosts, AIBOX, AI NVRs, and cluster servers. The Qualcomm QCS8550 SoC, with its built-in 48 TOPS NPU, offers exceptional edge computing capabilities and energy efficiency, providing reliable hardware support for high-performance IoT applications and enabling rapid deployment.
SoM: CORE-8550JD4
We've concentrated the system on chip (SoC) onto a tiny baseboard. Through various packaging methods, such as gold finger and board-to-board, this system on module can be flexibly configured in a variety of intelligent hardware. We also provide accompanying source code and reference design files, including schematics and PCB files, to facilitate customers to quickly build their own firmware and quickly implementing R&D projects based on the QCS8550.
Embedded PC: EC-A8550JD4 and AIBOX-8550
As for computer products, we offer a range of complete machine product forms, including the AIBOX and embedded host series, which effectively integrate software and hardware resources, enabling private deployment of mainstream AI large models.
At the same time, our products have powerful video processing capabilities, network communication capabilities, and support for multiple learning frameworks.
The firefly computers are tiny in size, rich in interfaces, and flexible in configuration. It has a wide temperature range and is stable, making it suitable for use in extreme industrial environments. This provides reliable hardware support and solutions for technology research and development across multiple fields.
Server: GS1-N2
In terms of servers, we provide cluster servers with standard 1U and 2U racks, which can integrate 10-72 computing nodes and provide up to 3456 TOPS computing power, providing powerful computing power for applications such as local deployment of large AI models and AI video analysis.
We have also developed an AI NVR server with a dual-core modular design. Its two integrated SoCs support independent operation and multi-tasking, enabling real-time decoding and analysis of multiple video streams.
Additionally, two built-in 3.5-inch SATA hard drive bays accommodate large video storage needs. Furthermore, the server can be configured with a BMC intelligent management system, providing a one-stop, deep learning development toolkit that significantly reduces development and maintenance costs.
Firefly's Qualcomm QCS8550 series products integrate low power consumption, maximum heterogeneous computing, Wi-Fi 7 connectivity, powerful image and video processing capabilities, and optimized AI architecture, and are widely used in autonomous mobile robots, drones, cameras, edge AI boxes, video transcoding, and cloud gaming.
