New Product Release: DX-U1200 Series Industrial Edge AI Computers with NVIDIA® Jetson Xavier NX™

  • Product & Service

Africa、Asia、Europe、Japan、Middle East、North America、Oceania、Latin America

The information included in this press release is current as of the publication date. The sales conditions, price, and specifications may be changed without notice.

Contec Co., Ltd. has developed an edge AI computer for industrial use, equipped with NVIDIA® Jetson Xavier NX™, and will begin receiving orders for the DX-U1200 series on July 27, 2021.

Simple dustproof model

Expansion slot model

Introducing the new high-end model of the DX-U1100 series with NVIDIA® Jetson Nano™, released in February 2021. Equipped with NVIDIA® Jetson Xavier NX™, this new product includes 384 Volta cores for the GPU, and two NVDLA (NVIDIA Deep Learning Accelerators) for improved deep learning performance. In ResNet-50 inference execution comparisons, the DX-U1200 enables faster edge inference execution for high-performance AI applications, including about 5.5 times faster FP16 arithmetic precision and about 21.5 times faster INT8 arithmetic precision.

Deep Learning Performance

ResNet-50 Inference Execution Resultsgraph

  • (Actual in-house measurement results from “Jetson_benchmarks” benchmark software; INT8 arithmetic precision not supported on the DX-U1100.)

Expansion slot model include a low-profile PCI Express slot for expanded functionality by means of Contec’s rich variety of low-profile PCI Express cards. Simple dustproof models also include interface connector caps and sheet metal pre-attached to prevent dust and dirt from entering the product vents.

As for the software environment, Ubuntu comes pre-installed just like the NVIDIA Jetson developer kit, and the NVIDIA JetPack™ SDK is available, so you can run applications created with the developer kit.


Model DX-U1200-3E0211 / DX-U1210-3E0211 DX-U1200P1-3E0211
Jetson Module Module NVIDIA® Jetson Xavier NX™
CPU 6 cores NVIDIA Carmel ARM® v8.2 64-bit CPU
GPU 384 cores NVIDIA Volta™ GPU 48 Tensor cores
DL Accelerator 2 x NVDLA (NVIDIA Deep Learning Accelerator) engine
Memory 8GB 128bit LPDDR4
Storage 16 GB eMMC
Software OS Ubuntu 18.04.5 LTS*1
SDK JetPack 4.4.1 [L4T 32.4.4]*1
Interfaces USB USB 3.2 GEN2(USB 3.1) Type A x 1, USB 2.0 Type A x 1, USB 2.0 microB x 1 (for debugging)
Display HDMI x 1, DisplayPort(eDP) x 1
LAN Gigabit Ethernet x 2
M.2*2 M.2 Key-E、M.2 Key-B
External Storage microSD memory card slot
Digital I/O DI x 4 (isolated), DO x 2 (isolated) (DI x 2 and DO x 2 are exclusive)
PCIe Slot - Low Profile PCI Express(x1) x 1
Others RTC built-in RTC built-in
Internal GPIO (for evaluation), MIPI CSI-2(for evaluation)
Simple Dustproof 〇(DX-U1210-3E0211) -
Physical Dimensions (mm)
(No bracket or projection included)
160(W)x92(D)x43(H) 160(W)x92(D)x75(H)
Rated Input Voltage 12 to 24VDC
Operating Temperature*3 -20 to 60°C (With PCIe board attached : -20 to 50°C) airflow 0.7m/s
Standard CE Marking (EMC Directive Class A, RoHS Directive), FCC Class A, VCCI Class A
  • *1
    More recent versions may be used without notice. Check the Contec website for more information.
  • *2
    Contact Contec for M.2 module installation.
  • *3
    Derating occurs due to the load conditions.
  • *
    NVIDIA, the NVIDIA Logo, and other NVIDIA marks are trademarks or registered trademarks of NVIDIA Corporation in the United States and other countries.

Related Links

Media Contact

Sales Promotion Group, Sales Division

Contact Us

  • *
    Company names and product names are generally trademarks or registered trademarks of each company.
  • *
    Please use the headquarters’ address if you wish to add an address in articles based on the content of this press release.

News Release

See all

To PageTop