[AutoDeploy]: Perf sweep SuperV3 nvfp4 with large workloads and identify perf gaps

[galagam]

🚀 The feature, motivation and pitch

Perf sweep with a large POR P0 workloads.

PRs that might be relevant:
#11515
#11624

Done criteria:

• Pareto curve comparing manual + autodeploy perf
• nsys traces
• Trace comparison report (AI generated, human validated)

See also: #11529

Alternatives

No response

Additional context

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.

[galagam]

See also PR #11581 which adds some prefill-specific optimizations