[Feature]: Support Nemotron 3 Super

### Feature Description

Submitting request, per:

```
2026-03-14 19:21:04 INFO __main__.py L586: start to quantize ./NVIDIA-Nemotron-3-Super-120B-A12B-BF16
2026-03-14 19:21:04 WARNING base.py L304: This MoE model has not been optimized by AutoRound yet, which may result in high RAM usage, Please consider submitting an issue to https://github.com/intel/auto-round/issues
```

### Motivation and Use Case

High RAM usage is bad, when the whole point is to quantize a model so that it fits smaller systems.

### Alternatives Considered

_No response_

### Definition of Done

_No response_

### Additional Context

_No response_