# Prefill-Decode Disaggregation Audit

## Description
Audits separate prefill and decode worker node topologies and calculates RDMA KV transfer latency vs GPU compute overlap.

## Category
`serving`

## Tags
`disaggregation`, `prefill-decode`, `rdma`, `split-serving`, `latency`

## When to Use
Evaluate disaggregated serving clusters where prefill and decode stages run on separate optimized GPU nodes.

## Operational Limitations
Evaluates RDMA network transfer bandwidth vs compute overlap bounds.

## Verification Script
Run the deterministic verification checker:
```bash
python3 scripts/check.py examples/synthetic.json
```
Exit codes:
- `0`: Check passed and thresholds satisfied.
- `1`: Threshold breach or operational violation detected.
- `2`: Schema error or invalid usage.
