MIC-743-AT: NVIDIA Jetson Thor AGX AI Inference System with 1TB Drive
MIC-743-ATStarting at $9,720.00
- Embedded with NVIDIA Jetson T5000 up to 2070 TFLOPS (FP4)
- Supports 1 x QSFP28 (4 x 25GbE)
- Supports 1 x 5GbE, 4 x USB 3.2 Gen 2, 1 x M.2 AEkey(WiFi), 1 x M.2 Bkey (LTE)
- Supports multiple purposes for generative AI applications using large AI models
Frequently Asked Questions
The MIC-743-AT provides:
- 1 x QSFP28 port (4 x 25GbE)
- 1 x RJ45 10/100/1000/2500/5000 Mbps Ethernet
- 4 x USB 3.2 Gen 2
- 1 x Micro USB OTG
- 1 x Micro USB console for debug (UART to USB)
- 1 x HDMI (max resolution 3840x2160 @ 60Hz)
- 1 x audio speaker out, 1 x MIC in on board
- 1 x Nano SIM slot on side of board
The unit accepts a DC power input of 19-36V through a 2-pin terminal block connector.
The MIC-743-AT operates from -10 to 35 °C with 0.7 m/s air flow. Operating humidity is rated at 95% at 40 °C (non-condensing), and it is rated for 3 Grms vibration from 5-500 Hz, random, 1 hour per axis.
The MIC-743-AT is certified to CB/UL/CE/FCC/BSMI/CCC. It does not carry RED certification.
The system includes 128GB LPDDR5X memory and a 1 x M.2 2280 M-key NVMe slot (PCIe x4). Ordering options list a pre-installed 1TB NVMe drive. Optional SQF PCIe/NVMe Gen.4 M.2 2280 OPAL drives in 2TB and 4TB capacities are also available.
The board includes 1 x M.2 2230 E-key slot (PCIe x1 + USB2.0, for WiFi), 1 x M.2 3052/3042 B-key slot (USB3.0 + USB2.0, for LTE), and an onboard TPM module.
The MIC-743-AT has a M.2 A/E-key slot for WiFi and a M.2 B-key slot for LTE, plus a Nano SIM card slot on the side of the board. An optional Wi-Fi 6 + BT module (EWM-W179M201E) with antennas is available.
The MIC-743-AT runs NVIDIA JetPack 7.0.
The MIC-743-AT measures 195 x 200 x 71.5 mm (W x D x H) including rubber feet, and weighs 2 kg. It is designed for desktop installation.
The MIC-743-AT is built around the NVIDIA Jetson T5000 module, featuring a 14-core Arm Neoverse V3AE (64-bit) SMP CPU with 64KB+64KB L1 cache per core, 1MB L2, and 16MB L3 cache. The GPU has 2560 NVIDIA CUDA cores and 96 5th Gen Tensor cores running at MAXN 1.57 GHz, with performance up to 2070 TFLOPS (FP4).
How-to Guides
- Locate the M.2 2280 M-key NVMe slot (PCIe x4) on the MIC-743-AT.
- Install a compatible M.2 2280 NVMe drive, such as the optional SQF PCIe/NVMe Gen.4 M.2 2280 OPAL drives (2TB or 4TB) listed in the ordering information.
- Note that units may ship with a 1TB NVMe drive pre-installed depending on part number ordered.
- For WiFi, use the M.2 2230 E-key slot (PCIe x1 + USB2.0). An optional Wi-Fi 6 + BT module (EWM-W179M201E) is available with antenna cables and dipole antennas.
- For LTE, use the M.2 3052/3042 B-key slot (USB3.0 + USB2.0).
- Insert a Nano SIM card into the SIM card slot on the side of the board if using an LTE module.
- Use the supplied 2-pin terminal block connector for power input.
- Connect a DC power source within the 19-36V input range to the terminal block.
- Optionally, use the included AC/DC adapter (100-240V input, 230W, 24V output) with an appropriate power cord for your region.
The MIC-743-AT utilizes a full sensor-to-inference pipeline, from multi-format camera input to real-time AI processing at the edge, all connected through high-speed ethernet with minimal latency. The system integration begins at the sensor level, supporting a variety of camera interfaces such as MIPI CSI and other high-speed industrial protocols.
Sensor data is routed through NVIDIA Holoscan Sensor Bridge, a sensor-to-ethernet streaming platform designed for low-latency applications, and aggregated via Advantech's high-speed sensor switch, which supports up to 2×10GbE uplinks depending on the system configuration. This data is then streamed directly into the MIC-743-AT, a compact, rugged, high-performance AI computing system built on the NVIDIA Jetson Thor module and powered by the NVIDIA Holoscan platform.
Leveraging camera over ethernet and high-throughput networking, the architecture ensures ultra-low latency and efficient data processing, which is crucial for safety-critical mobile and edge AI applications. The MIC-743-AT platform delivers an optimized, compact, and rugged solution for intelligent mobility, VLM (Vision Language Model) at the edge, and humanoid robotics.
| Product Available On | Oct 13, 2025 |
|---|---|
| Manufacturer | Advantech |
| Short Description | MIC-743 is an edge-ready platform built on the latest NVIDIA Jetson Thor, purpose-built for high-throughput AI workloads. It is ideal for robotics, automated inspection, mobility, and AI-based applications where real-time response and edge processing are mission-critical. |
| Processor | 14-Core Arm Neoverse V3AE (64-bit) SMP CPU Architecture |
| Graphics | 2560 NVIDIA CUDA Cores 96 5th Gen Tensor Cores MAXN: 1.57 GHz |
| Memory | 128GB LPDDR5X |
| Video Port | HDMI (Max. resolution 3840x2160 @ 60Hz) |
| Audio | 1 x audio speaker out, MIC in on board |
| M.2 | 1 x 2280 M-key NVMe (PCIex4) |
| Power Input Voltage | 19-36V |
| Power Type | Terminal Block 2 Pin |
| Dimension | 195 x 200 x 71.5 mm with rubber foot |
| Weight | 2kg |
| Operating Temperature | -10 ~ 35 °C with 0.7 m/s air flow |
| Humidity | 95% @ 40 °C (non-condensing) |
| Vibration | 3 Grms @ 5 ~ 500 Hz, random, 1 hr/axis |
| Max Resolution | 3840x2160 @ 60Hz |
| Display Type | HDMI |
| # of Port | 7 |