1. Understanding System-on-Module (SoM)

With the rapid advancement of semiconductor technology and the increasing integration of chips, the System-on-Module (SoM) has emerged as an ideal solution for the computational core design of electronic products. So, what is a System-on-Module (SoM)? A System-on-Module (SoM) is an embedded computing system that integrates key components, such as the central processing unit (CPU), memory, graphics processing unit (GPU), storage, expandable I/O interfaces, and peripherals onto a single printed circuit board (PCB). This compact yet highly integrated design delivers a modular computing core, offering users the flexibility to extend functionality by connecting external components via the available interfaces. As a result, SoMs provide customizable solutions to meet specific requirements.

2. Smartgiant’s SoM

Nexus is a typical SoM developed by Smartgiant, leveraging Xilinx Zynq UltraScale+ MPSoC FPGA technology. It seamlessly integrates the Xilinx Zynq UltraScale+ with multi-core ARM Cortex processors, extensive FPGA resources, and high-speed SerDes channels. The Nexus functions as a versatile controller, algorithm accelerator, and high-speed digital signal processor. Smartgiant offers end-to-end ODM services across the entire product lifecycle, including research and development, validation and mass production, as well as tailored solutions for hardware, logic, firmware, and software.

Smartgiant’s SoM -Nexus                                                                  Nexus Hardware Block Diagram

3. Advanced Cameras with SoM Technology

In the camera industry, system-on-module (SOM) solutions are increasingly addressing the growing demands of modern applications. Today’s cameras, whether used for industrial surveillance, home security or public facility monitoring, require advanced front-end recognition capabilities such as perimeter detection, object recognition, color identification, and motion detection. By integrating these features directly into the camera, the burden on back-end servers is reduced, leading to faster response times and improved system efficiency. The key to achieving this requires integrating a CPU, NPU, GPU, ISP, RAM, ROM, and communication units into the front end of the camera, creating a comprehensive computing system that is seamlessly embedded within the camera.

4. Overview of the RV1126 Core Board

The RV1126 core board, designed using SOM technology, is a high-performance embedded system-on-chip (SoC). It is  designed by Rockchip specifically for machine vision applications. It integrates a processor, an image signal processor (ISP), a neural network processing unit (NPU), and a variety of communication interfaces, supporting 4K video encoding and decoding with simultaneous encoding and decoding capabilities. The RV1126 is widely deployed in sectors such as smart security, video communication, edge computing, facial recognition, AI cameras, and video conferencing systems. It consolidates power supply, storage, and communication ports into the core board, with the IPC 38-Board standard structure design. With Smartgiant’s SDK, developers can easily develop a wide range of image-based applications on the core board, particularly AI recognition programs, enabling integrated front-end monitoring and recognition capabilities in camera solutions.

5. Smartgiant’s Horus AI Camera

The Horus AI Camera, developed by Smartgiant and built on the RV1126 core board, is a high-performance solution designed for AI-powered applications. Below are the key technological features of this product:

• AI Camera Module: Designed for AI applications,this camera module integrates the advanced RV1126 chip.
• Quad-Core ARM Cortex A7 Processor: Offers powerful computational performance for tasks requiring high concurrency and real-time processing.
• Memory Configuration: Supports DDR3 or DDR4 memory, providing high-bandwidth data access capabilities.

• 2.0 Tops AI Computing Power: Equipped with significant AI processing capabilities to efficiently run complex AI algorithms.

• Diverse Interface Options: Includes USB 2.0 x 1, Ethernet x 1, UART x 2, TF card expansion interface, HDMI interface, camera interface, and network interface, among others.
• IPC 38–Board Standard Structural Design: The camera’s enclosure can be quickly customized to meet various application scenarios.
• Video Codec: Compatible with multiple video codec standards, including H.264 and H.265, and capable of handling high-quality video streams, making it ideal for applications such as video surveillance and digital signage.
• Operating System Compatibility: Supports various operating systems, including Android and Linux, allowing developers to tailor applications for specific needs.

In summary, Horus is a customizable AI camera platform for software and hardware development. It offers a comprehensive SDK , which  includes system, driver, image processing, and AI application interfaces. It supports extensive development in Python, C and C++, thereby reducing both the time and expenses associated with software and hardware design and development. 

Below are some typical applications of the Horus AI Camera:

• Video Surveillance System: Equipped with advanced intelligent analysis algorithms and robust video processing capabilities, the Horus AI Camera enables features such as facial recognition and motion detection. This significantly enhances the intelligence of surveillance systems, making it ideal for smart monitoring solutions.

• Security Equipment: In security devices,the system module of the Horus AI camera can process video streams from multiple cameras, providing real-time monitoring and analytical capabilities. This functionality is crucial for enhancing the responsiveness and reliability of security systems.

• Industrial Automation:In the field of industrial automation, the Horus AI camera can be used for data acquisition, equipment monitoring, and control through the integrated RV1126 system module. With its high-performance computing power and real-time processing capabilities, the module enables intelligent management of industrial equipment, boosting productivity and ensuring greater safety.

From the SOM concept to the Horus AI camera developed by Smartgiant, our work highlights the seamless integration of advancements in electronic technology with market demands. With years of technical experience, Smartgiant has honed its expertise in designing and producing FPGA, CPU, and SoC systems. Additionally, we offer customized solutions tailored to specific needs. We welcome your inquiries and look forward to exploring opportunities for further discussions and collaborations.