Here's how the two Thor-based robotic controllers compare on the attributes you asked about:
| Attribute | EAR-100T (EverFocus) | AFE-A702 (Advantech) |
|---|---|---|
| GMSL camera inputs | 12x GMSL2 | 8x GMSL |
| Max Ethernet speed | 5GbE | 25GbE |
| Power input | 19–36 VDC | 20–28 V |
everfocus... +1everfocus.com ↗“12x GMSL2, 2x 5GbE”advantech.com ↗“io.max_ethernet_speed=25GbE”
Summary of the differences:
- GMSL inputs: The EAR-100T leads with 12 GMSL2 camera inputs vs the AFE-A702's 8 — more room to grow your camera array on the same controller. everfocus... +1everfocus.com ↗“12x GMSL2”advantech.com ↗“io.gmsl_camera_inputs=8”
- Ethernet speed: The AFE-A702 supports 25GbE, while the EAR-100T tops out at 5GbE — if you're streaming high-bandwidth sensor data off-board, the AFE-A702 has 5x the headroom on its network link. everfocus... +1everfocus.com ↗“io.max_ethernet_speed=5GbE”advantech.com ↗“io.max_ethernet_speed=25GbE”
- Power input: The EAR-100T accepts a wider input range (19–36 V), giving more flexibility with vehicle battery rails; the AFE-A702 is limited to 20–28 V. everfocus... +1everfocus.com ↗“power.power_input=19-36 V”advantech.com ↗“power.power_input=20-28 V”
Both are active full systems on Jetson Thor with the T5000 128GB module, 1035 TOPS (INT8 sparse) and 2070 FP4 TFLOPS — so your compute baseline is equivalent either way; the choice comes down to whether camera count or network bandwidth matters more for your robot. everfocus... +1everfocus.com ↗“compatibility.jetson_module=Jetson T5000 128GB”advantech.com ↗“Jetson T5000 128GB, Jetson T4000 64GB”