Specifications
Model
ARBOR AEC-6100 (SKUs AEC-6100-32 and AEC-6100-64)
Product Type
Rugged fan-cooled edge AI box PC
AI Module
NVIDIA Jetson AGX Orin
Module Option 1
AGX Orin 32GB — 8-core Arm Cortex-A78AE v8.2, 32 GB LPDDR5
Module Option 2
AGX Orin 64GB — 12-core Arm Cortex-A78AE v8.2, 64 GB LPDDR5
GPU
1792-core Ampere GPU with 56 Tensor cores (32GB) or 2048-core Ampere GPU with 64 Tensor cores (64GB)
Deep Learning Accelerators
2x NVDLA engines
Camera Interfaces
8x Fakra-Z connectors for GMSL 1/2 automotive cameras
Storage
2x M.2 Key M 2280 sockets on PCIe x4 plus 64 GB eMMC 5.1 and an external micro SD slot
LAN
2x 10/100/1000 Base-T Gigabit Ethernet on RJ-45
PoE Option
4x IEEE 802.3at (25.5 W / 48 V) Gigabit PoE+ on RJ-45 on the AEC-6200 variant
Serial
2x RS-232/422/485
CAN Bus
2x isolated CAN bus with CAN FD support
USB
1x USB 3.2 Gen2, 4x USB 3.1 (5 Gbps), 1x micro USB console, 1x micro USB OS flash
Digital I/O
16x isolated DIO — 8x DI and 8x DO
SIM
2x SIM card sockets
Expansion
1x PCIe Gen4 x8 slot, 1x M.2 Key B 3042/3052 (USB 3.0), 1x M.2 Key E 2230 (PCIe/USB)
Video Output
1x HDMI 2.1 up to 8K at 60 Hz
Video Encode
1x 4K60, 3x 4K30, 6x 1080p60 or 12x 1080p30 (HEVC)
Video Decode
1x 8K30, 2x 4K60, 4x 4K30, 9x 1080p60 or 18x 1080p30 (HEVC)
Operating Temperature
–25 to 70 °C at 30 W TDP; –25 to 55 °C at 40/50 W TDP
Shock / Vibration
Operating MIL-STD-810G Method 516.7 Procedure I and Method 514.6 Procedure I Category 4
Power Input
DC 9 V to 48 V on a 3-pin terminal block with frame ground and 16-mode software ignition control
Mechanical
Power, force-recovery and reset buttons; 6 antenna holes; wall, DIN rail and VESA mounting
Weight / Dimensions
3.8 kg; 260 x 182 x 69 mm
Operating System
Linux with NVIDIA JetPack SDK
Overview
The ARBOR AEC-6100 is a rugged edge AI computer built on the NVIDIA Jetson AGX Orin module, offered in 32 GB (eight Cortex-A78AE cores, 1792-core Ampere GPU with 56 Tensor cores) and 64 GB (twelve cores, 2048-core GPU with 64 Tensor cores) configurations. Both carry two NVDLA engines and run Linux with the NVIDIA JetPack SDK.
Camera and expansion capability are the headline differences from smaller Jetson boxes. Eight Fakra-Z connectors bring in GMSL 1/2 automotive cameras directly, an open PCIe Gen4 x8 slot accepts high-bandwidth add-in cards, and two M.2 Key M 2280 sockets on PCIe x4 provide NVMe storage alongside 64 GB of on-board eMMC 5.1 and an external micro SD slot. Video handling reaches 8K30 HEVC decode and 4K60 HEVC encode, with HDMI 2.1 output up to 8K at 60 Hz.
The platform is built for harsh deployment: –25 to 70 °C operation at 30 W TDP (–25 to 55 °C at 40/50 W), MIL-STD-810G operating shock and Category 4 vibration qualification, a 9–48 V DC input on a 3-pin terminal block with frame ground and 16-mode software ignition control, and wall, DIN rail or VESA mounting. Networking is handled by two Gigabit Ethernet ports, with two isolated CAN FD channels, 2x RS-232/422/485, 16 isolated digital I/O lines, two SIM sockets and six antenna holes completing the vehicle-class interface set.
Key Benefits
Eight FAKRA-Z GMSL 1/2 camera ports plus an open PCIe Gen4 x8 slot make it a multi-camera sensor-fusion platform rather than a single-stream box. AGX Orin 64GB with a 2048-core Ampere GPU and 64 Tensor cores provides top-tier Jetson inference capacity. 8K30 HEVC decode and HDMI 2.1 8K output cover high-resolution capture and display. MIL-STD-810G shock and vibration qualification with a –25 to 70 °C range suits rail, mining and off-highway vehicles. 2x isolated CAN FD, 16 isolated DIO, 2x SIM and 16-mode ignition control integrate directly with vehicle electrical systems.
Applications
Autonomous vehicle and ADAS development platforms, multi-camera sensor fusion, rail and off-highway machine vision, mining and construction equipment monitoring, defence and public-safety video analytics, smart-city intersection systems, and industrial inspection lines requiring PCIe expansion.
Request a Quote — ARBOR AEC-6100 — JETSON AGX ORIN BOX PC WITH 8X GMSL2, PCIE GEN4 X8 AND MIL-STD-810G
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →