Premio JCO-3000 NVIDIA Jetson Orin NX/Nano Rugged Edge AI Computers with 4x GbE

Many edge AI projects stall at the same point: a model that ran well on a developer kit now has to survive a vehicle, a factory wall or a roadside cabinet, with real cameras, real fieldbus and an unreliable power feed. The Premio JCO-3000 series closes that gap with rugged edge AI computers that carry the same NVIDIA Jetson Orin NX and Orin Nano Super modules in a fanless aluminium enclosure, adding the I/O and power handling that a developer kit leaves out.

Both JCO-3000 models accept any of four modules, so compute is chosen per project rather than per SKU: Orin NX 16GB at up to 100 TOPS, Orin NX 8GB at 70 TOPS, Orin Nano Super 8GB at 67 TOPS and Orin Nano Super 4GB at 34 TOPS. Storage is an NVMe SSD in an M.2 M-key slot running PCIe x4 (128 GB by default) plus a microSD slot. An M.2 B-key slot with an external dual nano SIM socket takes 4G or 5G, and an M.2 E-key slot takes Wi-Fi and Bluetooth. Fieldbus and control I/O are built in: CAN 2.0B, two RS-232/422/485 ports switched by the onboard MCU and 8 in/8 out isolated DIO, with an optional out-of-band management module on its own RJ45 port.

The range differs only in how the four GbE ports are used. The JCO-3000-ORN-B-4L-P runs four RJ45 GbE ports without PoE, for segmented networks, self-powered cameras and controllers. The JCO-3000-ORN-B-4P-P adds a 120 W PoE+ module so each of the four ports can deliver up to 25 W to IP cameras or sensors, with a total PoE budget of up to 60 W on a 12 V supply and up to 120 W on 24 V.

Typical deployments include autonomous mobile robots and AGVs, inspection cells with several GigE cameras, vehicle and fleet analytics, and traffic or parking enforcement. Every JCO-3000 model takes 9 to 36 V DC on a 3-pin terminal block with adjustable ignition management, operates from -20 °C to 60 °C at 15 W or -20 °C to 55 °C at 25 W, and is listed with UL 62368-1, CE, FCC Class A and E-Mark (E24). Shock and vibration are designed to comply with MIL-STD-810H, and the 192 x 140 x 58 mm enclosure wall mounts as standard, with a DIN rail option. These edge AI computers run Ubuntu with NVIDIA JetPack.