• 边缘计算(英).png
边缘计算(英).png
+

TME-2-SH Edge Computing Device

TME-2-SH is developed and designed based on the domestically produced ShanHai module, delivering a computing power of 200/240 TOPS. It features a domestically produced 8-core 64-bit Arm architecture CPU and a domestic GPU, equipped with onboard 16/32 GB HBM2E graphics memory. The device supports external SATA storage and offers abundant I/O resources. Powered by DC 9–16V, it is suitable for industry applications such as intelligent transportation, machine vision, smart security, and smart logistics, effectively meeting the demands for independent control and domestic innovation.
  • Commodity name: TME-2-SH Edge Computing Device

Category:

Product Inquiries:

  • Product Description
  • TME-2-SH is developed and designed based on the domestically produced ShanHai module, delivering a computing power of 200/240 TOPS. It features a domestically produced 8-core 64-bit Arm-based CPU and an integrated domestic GPU. The board comes equipped with 16/32 GB of HBM2E graphics memory and supports external SATA storage. With abundant I/O resources and a DC power supply range of 9–16 V, TME-2-SH is well-suited for industry applications such as intelligent transportation, machine vision, smart security, and smart logistics, effectively meeting the demands for independent and controllable solutions.

     

    Features:

    1. Developed and designed based on the domestically produced ShanHai module, with computing power of 200/240 TOPS, delivering performance comparable to NVIDIA AGX ORIN.

    2. Equipped with a domestically produced 8-core ARM 64-bit CPU;

    3. Equipped with a domestically produced GPU featuring 4,096 NVIDIA CUDA® Core cores and 512 Tensor Core cores;

    4. 1*M.2 2280 M-Key NVMe, up to 8TB

    5. 1*HDMI port, 1*debug interface, 2*USB 3.0 ports, 1*RS485 port, 1*RS232 port

    6. 1*10-gigabit Ethernet port (PTP supported), 1*Gigabit Ethernet port (PTP supported);

    7. SATA 3.0 interface for external hard drive connection;

    8. The CUDA ecosystem is compatible with mainstream AI frameworks and models, enabling seamless migration of business software without any noticeable impact.

    9. Aluminum rectangular profile housing with fanless passive cooling design;

    10. The entire machine features an IP67 protection rating, making it suitable for a variety of applications including transportation, industrial settings, and security scenarios.

     

    Technical Specifications:

    Indicator Item

    Parameter Value

    GPU computing power

    16 TFLOPS @ FP32 (Tensor Cores) | 8 TFLOPS @ FP32 (CUDA Cores)
    64 TFLOPs @ FP16, 200 TOPS @ INT8

    CPU Performance

    8-core 64-bit, with a main frequency of 1.6 GHz

    Memory

    16GB of GPU memory + 16GB of dedicated CPU cache

    32GB GPU Memory + 16GB CPU Independent Cache

    Video Encoding

    16 x 1080P 30-channel parallel encoding

    Video Decoding

    32 x 1080P 30-channel parallel decoding

    Ethernet

    Supports 1 x 1000 Mbps and 1 x 10,000 Mbps

    Storage Capacity

    1* M.2 2280 M-Key, compatible with NVMe high-speed storage cards

    Indicator light

    1* System Status Indicator Light, 1* Gigabit Network Indicator Light, 1* 10-Gigabit Network Indicator Light, 1* Hard Drive Light

    Extended Interface

    1*RS485, 1*RS232, 1*GPS, 2*USB 3.0, 1*HDMI, 1*Upgrade Interface, 1*Debug

    Power Requirements

    DC 12V

    Operating System

    Ubuntu/Galaxy Kirin, etc.

    Dimensions

    282mm × 192mm × 75mm

    Operating Temperature

    Module Operating Temperature: -40℃ to 70℃

    Power Consumption

    80W

Keywords:

Obtain Quotations and Materials

Note: Please leave your contact information, and our professional team will get in touch with you as soon as possible!