add benchmark data

[vMaroon]

Summary

Adding a new benchmark data, used for an llm-d blog.

[Copilot]

Pull Request Overview

Adds comprehensive benchmark data for an inference performance study comparing different scheduling strategies (precise, estimated, load, and random) for LLM workloads, including detailed metrics, configurations, and visualizations for a blog post.

• Adds a complete benchmark report with workload configuration, scheduler configurations, and performance metrics
• Includes performance comparisons across different QPS levels and scheduling strategies
• Provides per-pod EPP metrics and visualization references for analysis

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[nilig]

/lgtm

[vMaroon]

/approve