Debashis just wrote about the Firefly EC-ThorT5000 embedded PC powered by an NVIDIA Jetson Thor T5000 system-on-module yesterday, and the ASUS IoT PE3000N is another rugged, modular Edge Platform powered by the 2,070 FP4 TFLOPS SoM offering MIL-STD-810H compliance, 12-60V input, 25GbE links, and support for up to sixteen GMSL cameras.
The PE3000N features one HDMI 2.0 video output, GbE and 10GbE RJ45 ports, M.2 sockets for storage and wireless expansion, a PCIe x4/x8 slot, and several USB ports, as well as four 25GbE links and AGX CSI connector for up to 16x GMSL cameras for “high-bandwidth sensor fusion and advanced machine vision”, even in the most challenging environments. The Jetson T5000 embedded system also offers a scalable chassis and modular I/O layers that support interfaces such as PoE, GMSL, CAN, and QSFP28 within a compact 2U height.
ASUS IoT PE3000N specifications:
- SoM – NVIDIA Jetson Thor T5000 module
- CPU – 14-core 64-bit ARM Neoverse-V3AE @ up to 2.6GHz
- GPU – 2560-core Blackwell GPU with 96 Tensor Cores, 10-TPC MIG
- VPU
- Video Decode – 4x 8Kp30 (H.265), 10x 4Kp60 (H.265), 22x 4Kp30 (H.265), 46x 1080p60 (H.265), 92x 1080p30 (H.265), 82x 1080p30 (H.264), 4x 4Kp60 (H.264)
- Video Encode – 6x 4Kp60 (H.265), 12x 4Kp30 (H.265), 24x 1080p60 (H.265), 50x 1080p30 (H.265), 48x 1080p30 (H.264), 6x 4Kp60 (H.264)
- AI performance – 2070 TFLOPS FP4 (Sparse)
- Memory – 128 GB 256-bit LPDDR5X 273 GB/s
- Video Output – HDMI 2.0
- Camera
- 16-lane MIPI CSI via AGX CSI connector (for GMSL)
- Fakra connectors via stackable unit
- Audio
- 3.5mm audio input jack
- 3.5mm audio output jack
- Networking
- Gigabit Ethernet RJ45 port
- 10GbE RJ45 port
- Additional Ethernet ports via stackable unit
- Optional WiFi and Bluetooth via M.2 socket
- Optional 4G LTE/5G cellular via M.2 socket and nano-SIM slot
- USB
- 4x USB 3.2 Gen2 Type-A ports
- USB 3.2 Gen2 Type-C port
- USB 3.2 Gen2 OTG Type-C port (OS Flash)
- 2x USB 2.0 via headers
- Serial
- 2x RS-232/422/485
- 2x CAN bus
- Debug Port – USB Type-C port for debug console
- Expansion
- M.2 M key 2242/2260/2280 socket for M.2 SSD, or other M.2 PCIe modules
- M.2 E key 2230 socket for Wi-Fi/BT, etc…
- M.2 B key 3042/3052 for 4G/5G, etc…
- PCIe x4/x8 slot
- 4x 25GbE for 1/5/10/25Gbps cable or fiber ethernet, e.g. RJ45, M12, QSFP28 100G, etc
- UART, I2C, SPI, Control Pin, GPIO
- 2×5 DI & DO
- Security – On-board TPM v2.0
- Misc
- Time Sync with PPS, PTP, Camera Sync
- Watchdog timer
- Wall-mount (support assembly with DIN rail clips)
- Power Supply – 12 to 60 VDC, 3-pin terminal block (1-pin for remote power button)
- Dimensions
- Board – 145 x 145 mm
- System
- 165 x 165 x 68 mm (main sku)
- 165 x 165 x 97 mm with stackable unit (within 2U height)
- Weight – TBD
- Temperature Range – Operating: -20 to 60°C; storage: -40 ~ 85°C
- Relative Humidity – 10 ~ 95%, non-condensing
- Vibration – MIL-STD 810H Method 514.8, Category 4 (Truck transportation over US highways); 5 Grms, 5-500 Hz, 3 axes
- Shock – MIL-STD 810H Method 516.8, Procedure I; 40 Grms/11 ms
Like other NVIDIA Jetson platforms, it is supported by the Jetpack SDK with Ubuntu Desktop by default. It provides plenty of computational power for generative AI models, such as visual language models (VLMs) and large language models (LLMs), as well as real-time video analytics and autonomous control tasks. Supported frameworks include NVIDIA Isaac for robotics, NVIDIA Holoscan platform for real-time sensor processing, and NVIDIA agentic AI workflows such as NVIDIA Blueprint for Video Search and Summarization (VSS) from NVIDIA Metropolis.
The ASUS IoT PE3000N is more flexible than the Firefly embedded PC we just covered, thanks to a modular design that allows users to add stackable units underneath the main unit. Two standard stacks are available: one with six GbE ports with PoE, and one with four 10GbE M12 ports and two Fakra connectors for up to 8 GSML cameras. Customized stacks in the same form factor are also possible.
The ASUS IoT PE3000N will officially launch in Q1 2026, and pricing information may become available at that time. Additional information may be found on the product page and the press release.
Thanks to TLS for the tip.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.
Support CNX Software! Donate via cryptocurrencies, become a Patron on Patreon, or purchase goods on Amazon or Aliexpress. We also use affiliate links in articles to earn commissions if you make a purchase after clicking on those links.
