How Cloudflare uses the world’s greatest collection of performance data to make the world’s fastest global network even faster

This post details how Cloudflare is leveraging its extensive network data and Real User Measurement (RUM) data to dynamically optimize congestion control algorithms. It introduces the concept of predicting RUM measures from passive logs and using this to fine-tune congestion control parameters, leading to an average of 10% performance improvement. The post also discusses the migration to a Rust-based network technology stack for greater flexibility in experimenting with congestion control algorithms.