AAEON BOXER-8742AI Packs Jetson T4000 Power in a Fanless Edge Box PC

BOXER-8742AI Fanless Embedded Box PC

The AAEON BOXER-8742AI is a compact, fanless edge AI embedded box PC designed for modern industrial and intelligent infrastructure deployments. Built with a modular hardware design, it addresses demanding edge workloads such as real-time AI inference, machine vision, robotics, and automated decision-making applications. Despite its small footprint, the system delivers high compute performance and is based on the NVIDIA Jetson T4000 platform for edge AI processing.

Compared to the BOXER-8741AI, the BOXER-8742AI integrates a 12-core Arm Neoverse-V3AE 64-bit CPU alongside a 1,536-core NVIDIA Blackwell GPU with 64 Tensor Cores. This hardware combination delivers up to 1200 TFLOPS of AI performance, enabling support for large and complex AI models used in smart factory automation, autonomous mobile robots (AMRs), industrial inspection systems, and intelligent transportation applications.

Previously, we covered several embedded box PCs such as the BOXER-8260AI, BOXER-8261AI, BOXER-8170AI, and Vecow SPC-7000/7100. Feel free to check those articles if you’re interested in similar products.

BOXER-8742AI Fanless Embedded Box PC Rear View

BOXER-8742AI Specifications:

• AI accelerator: NVIDIA Jetson T4000 (Jetson Thor platform)
• CPU: 12-core Arm Neoverse-V3AE 64-bit processor
• GPU: 1,536-core NVIDIA Blackwell GPU with 64 fifth-gen Tensor Cores
• AI performance: Up to 1200 TFLOPS
• Memory: 64GB 256-bit LPDDR5X
• Storage: 1× M.2 2280 M-Key (NVMe SSD)
• Display: 1× HDMI 2.0
• Ethernet: 1× 5GbE RJ45, 4× 1GbE RJ45 with PoE (802.3af/at, 60W max)
• USB: 4× USB 3.2 Gen 2 Type-A
• Serial and field I/O:
  • 1× DB-15 (RS-232/422/485 + 8× DIO)
  • 1× DB-9 (RS-232 + 2× CAN FD)
• Wireless expansion:
  • 1× M.2 3042/3052 B-Key (4G/5G, USB 3.0 + PCIe)
  • 1× M.2 2230 E-Key (Wi-Fi / Bluetooth / NVMe)
• SIM: 2× SIM slots
• Antenna: 6× external SMA connectors
• Expansion (storage): 1× M.2 2280 M-Key (NVMe)
• Misc:
  • Recovery Button x 1
  • Reset Button x 1
  • Power Button x 1
• Power input: 12–24V DC via 2-pin Phoenix connector
• Dimensions: 210 × 164.2 × 74 mm
• Operating temperature: −25°C to 60°C (with 0.7 m/s airflow)
• Storage Temperature: -40°F – 185°F (-40°C ~ 85°C), non operation
• Storage Humidity 5%~ 95% @ 40°C, non-condensing
• Anti-Vibration: 3.5G / 5~500Hz / operating
• Anti-Shock: 50G peak acceleration ( 11 msec.)
• Certifications: CE, FCC Class A

Dimensions

Despite integrating a wide range of high-end features, the BOXER-8742AI adopts a fully fanless design, allowing it to operate reliably in harsh industrial environments while occupying minimal installation space. Its compact enclosure makes it suitable for deployment in space-constrained factory floors and edge installations. The system supports NVIDIA JetPack 7.0 and later, simplifying the development and deployment of AI applications using CUDA, TensorRT, and container-based workflows.

At the time of writing, AAEON has not officially announced pricing for the BOXER-8742AI, as many of its high-end edge AI platforms are offered on a quote-only basis. Based on recent announcements, the system is expected to enter commercial availability around mid-2026. While pricing has not been confirmed, industry estimates suggest it may fall in the $3,000 range, with final costs varying by configuration and region. Interested customers should contact AAEON or an authorized distributor for accurate pricing and lead-time details.

Images used courtesy of AAEON