Specifications
Model
Advantech MIC-741-AT
Platform
NVIDIA Jetson Thor (supports Jetson T5000 and Jetson T4000 modules)
AI performance
Up to 2070 TFLOPS (FP4)
Networking
1x SFP28 (25GbE)
Ethernet
2x 1GbE + 2x 10GbE
USB
6x USB 3.2 Gen 2
Storage expansion
2x M.2 2280
Wireless expansion
1x M.2 E-Key (Wi-Fi); 1x M.2 B-Key (LTE / 5G)
Fieldbus
2x isolated CAN FD
Target workloads
High-throughput edge AI inference, robotics, physical AI
Series
Advantech MIC-AI inference systems
Use case
SATA/SAS-free edge inference node for demanding AI workloads
Overview
The Advantech MIC-741-AT is a high-performance edge AI inference system built on the NVIDIA Jetson Thor platform, supporting the Jetson T5000 and T4000 modules and delivering up to 2070 TFLOPS at FP4 precision. It targets demanding artificial-intelligence workloads that previously needed a rack-level GPU server, in an edge-optimised form factor.
I/O is designed for physical-AI and robotics deployments: an SFP28 25 Gigabit Ethernet port, dual 1 GbE and dual 10 GbE links, six USB 3.2 Gen 2 ports, two M.2 2280 storage slots, M.2 E-Key and B-Key for Wi-Fi and LTE/5G, and two isolated CAN FD interfaces for direct connection to vehicle and machine buses. It sits in Advantech's MIC-AI series of Jetson-based inference systems.
Key Benefits
Jetson Thor, up to 2070 TFLOPS FP4 brings data-centre-class inference to the edge. SFP28 25GbE plus 2x 10GbE handle high-bandwidth sensor and camera feeds. 2x isolated CAN FD connects directly to robotics and vehicle buses. Dual M.2 2280 and 5G/Wi-Fi slots cover storage and connectivity.
Applications
Physical AI and humanoid robotics, high-throughput machine vision, autonomous systems, and industrial edge inference where low latency and high TOPS are both required.
Request a Quote — ADVANTECH MIC-741-AT — JETSON THOR EDGE AI INFERENCE SYSTEM, UP TO 2070 TFLOPS (FP4)
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →