Reproduce “Autobahn: Seamless high speed BFT” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Neil Giridharan, Florian Suri-Payer, Ittai Abraham, Lorenzo Alvisi, Natacha Crooks
Autobahn introduces a Byzantine Fault Tolerant (BFT) protocol that simultaneously achieves low latency, high throughput, and robust recovery from transient network disruptions ("blips"). It demonstrates that the typical trade-off between these properties in existing BFT protocols is not fundamental, outperforming both traditional and DAG-based BFTs in specific metrics while avoiding "protocol-induced hangovers".
Key insights
- The trade-off between low latency and robust recovery in BFT protocols is not fundamental and can be overcome by optimally decoupling data dissemination from consensus.
- Real-world BFT performance requires new evaluation metrics like "hangovers" and "seamless partial synchrony" to account for intermittent network disruptions.
- A two-layer architecture, where an asynchronous data layer continuously processes transactions and a fast synchronous layer orders "snapshots" of this data, enables superior performance across diverse network conditions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.