AEC-8000 Series
AEC-8000 Series — High‑Performance Edge AI System with NVIDIA Jetson Thor T5000
- 2 x GbE RJ-45
- 1 x QSFP for 4x25GbE
- 1 x M.2. E key for Wi-Fi 6E
- 1 x M.2 B key for 4G/ 5G
- 2 x M.2. M key for SSD
- GMSL support (AEC-8100)
- PoE support (AEC-8200)
- 8 x USB 3.2 Type A support (AEC-8300)
Extreme Performance Edge AI System Powered by NVIDIA® Jetson Thor™
The AEC-8000 Series is a next generation industrial Edge AI computer built to meet the most demanding real time AI workloads at the edge. Powered by NVIDIA® Jetson Thor™ T5000 with Blackwell GPU architecture, AEC-8000 delivers data center class AI performance in a compact, industrial grade platform, enabling instant decision making where latency, bandwidth, and reliability matter most. Designed for smart transportation, multi camera vision AI, autonomous systems, and AI driven infrastructure, AEC-8000 transforms raw data into actionable intelligence directly at the edge.
Data Center Class AI Performance at the Edge
At the core of AEC-8000 is a powerful combination of a 14 core Arm® Neoverse™ V3AE CPU and an NVIDIA Blackwell GPU with 96 Tensor Cores, delivering up to 2070 TFLOPS (FP4) of AI compute performance. This enables simultaneous execution of multiple AI models, real time inference, and high throughput data processing, without relying on cloud resources.
AEC-8000 is ideal for:
- Multi stream AI video analytics
- Real time traffic and event detection
- Autonomous perception and sensor fusion
- High density AI inference at the industrial edge
Built for Massive Video and Sensor Data Throughput
To eliminate bandwidth bottlenecks in data intensive environments, AEC 8000 integrates a QSFP interface supporting 4 × 25GbE, enabling ultra high speed data ingestion and aggregation. This architecture supports over 64 HD video streams for simultaneous analysis, making it a powerful platform for vision centric AI deployments.
Multiple system variants allow optimal configuration for different use cases:
- AEC-8100: 8 × GMSL camera connectors for automotive grade vision systems
- AEC-8200: 8 × GbE PoE ports for rapid IP camera deployment
- AEC-8300: Expanded 8 x USB 3.2 support for additional AI peripherals
- AEC-8400: 8 × GMSL camera connectors for automotive grade vision systems & 8 × GbE PoE for rapid IP camera deployment
- AEC-8500: Expanded USB 3.2 support for additional AI peripherals & 8 × GMSL camera connectors for automotive grade vision systems
This modular design provides maximum flexibility while reducing integration complexity.
Industrial Grade Design for In Vehicle and Harsh Environments
AEC-8000 is engineered for continuous operation in demanding edge environments. With 9–36V wide range DC input, ACC / IGN ignition control, and rich industrial I/O—including CAN FD, GPIO, and RS 232/422/485—the system is well suited for both fixed and mobile deployments.
Typical deployment scenarios include:
- Smart transportation and roadside AI systems
- In vehicle vision AI and fleet intelligence
- Industrial automation and outdoor AI nodes
Robust thermal design and industrial grade components ensure stable, long term operation under sustained AI workloads.
Scalable Edge AI Platform for Long Term AI Evolution
Beyond raw performance, AEC-8000 is designed as a future proof Edge AI platform. Support for Ubuntu and NVIDIA JetPack™, along with PCIe Gen5 NVMe storage, Wi Fi 6E, and 5G expansion, ensures seamless integration with modern AI development workflows and evolving edge architectures.
From today’s real time inference needs to tomorrow’s large scale AI models, AEC-8000 provides a solid foundation for scalable and sustainable Edge AI deployments.
System |
|
| CPU | 14-core Arm® Neoverse™ V3AE 64-bit CPU |
|---|---|
| Memory | 128GB LPDDR5X DRAM |
| GPU | 2,650-core NVIDIA Blackwell™ GPU with 5th GEN 96 Tensor Cores |
| Storage | 1x NVMe M.2 Key M 2280 x2 PCIe Gen5 (8xPoE, 8xUSB
board, either one) 1x NVMe M.2 Key M 2280 x4 PCIe Gen5 |
| TPM | SLB9672XU2.0 |
I/O Interface |
|
| Serial Port | 2 x RS-232/422/485 ports with DB9 connector |
| LAN | 2 x GbE ports w/ RJ-45 connector, one port is NCSI
OOB on board (Optional) 1 x QSFP for 4 x25GbE |
| USB Port | 4 x USB 3.2 Type A 1 x USB 3.2 Type C 1 x USB 3.2 Type C for BSP install (OTG) Optional 8 x USB 3.2 Type A (via daughter board, AEC- 8300) |
| Expansion | 2x10-pin Euro Terminal block (1x CAN-FD (Isolated 3KV),
3x CAN-FD with Transceiver, +5V output 1A , +3.3V output
1A,UART, I2C) 1 x M.2 B key (3042 for 4G LTE module or 3052 for 5G module) 1 x M.2 E key (for Wi-Fi 6E) 2 x M.2 M key (for SSD) 2 x SIM sockets |
| GPIO | 8-DI/ 8-DO with DB9 connector |
| Video Port | 1 x USB Type C DP ALT mode 2 x HDMI output |
| GMSL | 8 x Fakra-Z connectors for GMSL automotive cameras (AEC-8100) |
| PoE | 8 x GbE PoE + LAN w/ RJ-45 connector (AEC-8200) |
| Audio | MIC in/ Line out with 3.5mm phone jack |
Power Requirement |
|
| Power Input | DC 9V ~ 36V with 3-pin terminal block, AT/ ATX Mode
selectable ACC IN/ IGN control (ACC optional, via switch) |
|---|---|
| Thermal solution | 1 x SOM Fan 2 x Chassis fan (12V fan wafer) |
Mechanical |
|
| Button | 1 x Power Button, 1 x Force Recovery Button |
| Antenna | 4 x Antenna holes for WiFi |
| Mounting | Wall-mount |
| LED indicator | 1 x System power 1 x Input power |
| Weight | 4Kg |
| Dimensions (W x D x H) |
211 x 194 x 111mm |
OS Support |
|
| System | Ubuntu/ NVIDIA JetPack 7.0 |
Ordering Information |
|
| AEC-8000 (Main SKU) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS |
| AEC-8100 (BTO) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS, w/ 8 x GMSL 1/2 |
| AEC-8200 (BTO) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS, w/ 8 x PoE |
| AEC-8300 (BTO) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS, w/ additional 8 x USB 3.2 |
| AEC-8400 (BTO) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS, w/ 8 x PoE & 8 x GMSL 1/2 |
| AEC-8500 (BTO) | Jetson Thor T5000 128GB BOX PC, up to 2070 TFLOPS, w/ additional 8 x USB 3.2 & 8 x GMSL 1/2 |
| Description | Size | Type | Version | Date | Download |
|---|---|---|---|---|---|
|
Description
|
Size
-
|
Type
pdf
|
Version
|
Date
2026-03-18
|
Download
|
