# Bounded Work Queue vs. Thread-Per-Request

These are two different concurrency strategies for handling incoming work, and they trade off differently across throughput, latency, resource usage, and failure modes.

## Thread-per-request

Each incoming request spawns a new thread that runs to completion, then dies.

```
