Neousys Nuvo-9154GC Intel 14th/13th/12th Gen Core AMR Computer with 4x GMSL2 Cameras and 150W NVIDIA RTX GPU

An autonomous mobile robot is only as reliable as the timing between its cameras, LiDAR and motion controller. The Neousys Nuvo-9154GC is an AMR computer for warehousing, logistics and agriculture, combining Intel 14th/13th/12th Gen Core processors with a 150 W NVIDIA RTX GPU, four GMSL2 camera inputs on one Mini-FAKRA connector, and GPS PPS input and output for frame synchronisation.

The LGA1700 socket on the Intel Q670E chipset takes 14th Gen Core i3 to i9 and 13th and 12th Gen Core E-series processors at 35 W or 65 W, up to 24 cores and 32 threads, with up to 128 GB of DDR5-4800 for SLAM and LiDAR-based path planning on the CPU. The GPU sits in a PCIe x16 Gen3 cassette slot with a dedicated bracket against shock, and with a 150 W card the Nuvo-9154GC delivers up to 19.2 TFLOPS of FP32 for object recognition and contour analysis.

The GMSL2 input supports configurations from four 2 MP cameras at 60 fps to two 8 MP plus one 3 MP at 30 fps, and Neousys provides driver-ready support under Linux with a porting mode for unlisted cameras. Five 2.5GbE ports and one Gigabit port carry screw-lock retention, with PoE+ on ports 3 to 6 from a 100 W budget. A 20 Gbps USB-C port, six USB 3.2 Type-A ports and four COM ports complete the I/O, storage pairs an M.2 Gen4 x4 NVMe socket with two 2.5-inch SATA drives in RAID 0 or 1, and mini-PCIe and M.2 B-key sockets take Wi-Fi, 4G and 5G.

Power is an 8 to 48 V DC terminal block with built-in ignition control and 15 pre-defined on and off delay modes on a rotary switch, so the AMR computer follows the robot’s power state. With a 35 W processor the Nuvo-9154GC operates from -25 °C to 60 °C, or to 50 °C configured at 65 W, with MIL-STD-810H vibration and shock, at 240 x 225 x 110.5 mm and 3.97 kg. Where the cameras are GigE rather than GMSL2, the Nuvo-9160GC offers the same compute class without the FAKRA input.