Vecow EAC-7200F-T5000 NVIDIA Jetson Thor T5000 Robot Computer with 16x GMSL, 10GbE and CAN-FD

Surround-view perception on an autonomous vehicle or humanoid robot can mean a dozen or more cameras feeding one computer. The EAC-7200F-T5000 is built for that load: the 16-camera robot computer in the Vecow EAC-7000 series, with the NVIDIA Jetson Thor T5000, two 10GbE ports and four isolated CAN-FD ports.

The NVIDIA Jetson Thor T5000 in the EAC-7200F-T5000 combines a 14-core Arm Neoverse V3AE CPU with a 2,650-core Blackwell GPU and 96 fifth-generation Tensor cores, on 128 GB of 256-bit LPDDR5X. Vecow quotes the Thor range at up to 2,070 FP4 TFLOPS, which leaves headroom for perception, planning and large vision-language models on the same computer. It runs Ubuntu 24.04 with JetPack 7.0 and above, and supports NVIDIA Isaac Perceptor on Isaac ROS and the Holoscan Sensor Bridge for low-latency sensor pipelines.

Networking on the EAC-7200F-T5000 is two 2.5GbE ports and two 10GbE ports, and every Ethernet port can be ordered with X-coded M12 connectors. Four x4 mini-Fakra connectors take 16 GMSL 1/2 automotive cameras. Field I/O covers four isolated CAN-FD ports, 16 isolated DIO, two RS-232/422/485 ports and four USB 3.1 ports, with two HDMI outputs up to 8K. M.2 Key B and Key E sockets with two nano SIM slots and eight antenna positions take 5G, 4G and Wi-Fi, and an M.2 Key M PCIe x4 slot holds the NVMe drive.

The 260 x 215 x 88 mm chassis runs on 9-50 V DC with 16-mode software ignition control and is rated to MIL-STD-810G operating shock and vibration, with EN 50155 EMC listings. In 77 W mode it runs fanless from -20 °C to 45 °C, or to 70 °C with the fan sink. For eight cameras rather than sixteen, see the EAC-7100F-T5000.