Get a head start designing your imaging AI applications with new compact smart AI module

SIMCom has announced two new smart modules containing AI and all the other features needed to easily and quickly add AI imaging capabilities to a wide range of new applications.

Mats Fischer, SIMCom Europe Sales Director, commented: “Our new smart AI module targets emerging imaging IoT applications that require AI capabilities for positioning and identification. By incorporating the AI ​​core, MCU, and connectivity capabilities such as Wi-Fi and Bluetooth into one compact module, design engineers get a head start and can run easily available software.”

SIM8666 and SIM8668 series are the company’s basic Android system LTE smart AI modules. Available in a compact 43mm x 44mm x 3.2mm package, the device features Rockchip’s 64-bit quad-core RK3566, RK3568 IoT chip, ARM Mali-G52 GPU, 8M ISP HDR, and 1T NPU computing utility. The module supports Wi-Fi and BT short-range communication, H.264/H.265 4k@60fps decoding and H.264/H.265 1080p@60fps encoding, JPEG encode/decode, etc.

Series modules are equipped with a rich set of interfaces and can be connected to devices such as cameras, display screens, audio, sensors, etc. The device supports multidimensional data collection and man-machine interaction. Interfaces such as LVDS, MIPI-DSI, RGB, eDp, HDMI2.0, EBC, CSI, UART, SPI, I2C, GPIO, PWM, USB, PCIe, and SATA make the smart AI module extremely versatile. The embedded NPU supports INT8/INT16 mixed operations and facilitates the conversion of algorithmic models based on various frameworks such as TensorFlow, MXNet, PyTorch, and Caffe, making it suitable for modules in applications such as IoT gateways, smart business displays, industrial controls, smart hospitals, smart central control, and smart security.

Fischer concludes: “Our new smart AI module enables our customers to rapidly develop and implement high-computing, multimedia, lightweight AI designs for never-before-seen applications at low cost.”