Modern Multi-proposer consensus implementations

Multi-proposer consensus protocols let multiple validators propose blocks in parallel, breaking the single-leader throughput bottleneck of classic designs. Yet the modern multi-proposer consensus implementation has grown a lot since HotStuff. This workshop explored the implementation details of recent advances – DAG-based approaches like Narwhal and Sui’s Mysticeti – and revealed how implementation details translate to real-world performance gains. We focused on the nitty-gritty: how network communication patterns and data handling affect throughput and latency. New techniques such as Turbine-like block propagation (inspired by Solana’s erasure-coded broadcast) and lazy push gossip broadcasting dramatically cut communication overhead. These optimizations aren’t just theoretical – they enable modern blockchains to process over 100,000 transactions per second with finality in mere milliseconds, redefining what is possible in decentralized systems.

Session highlights:

• Evolution of Multi-Proposer Consensus: How we moved from single-leader BFT protocols to parallel proposal systems.
• Traditional vs. Consistent Broadcast Approaches: A comparison between classical consensus and Byzantine Consistent Broadcast models used in projects like Linera and Sui, and more recently delta and pod. Discussion of how consistent broadcast ensures all nodes receive the same transactions (often with less coordination), and the trade-offs in complexity, finality guarantees, and network assumptions.
• Network-Layer Optimizations: How advanced networking techniques amplify consensus performance. Breakdown of gossip-based communication using epidemic broadcast trees (Plumtree), and Solana’s Turbine protocol, a tree-structured broadcast that splits blocks into smaller chunks and disperses them with error-coding. These approaches drastically reduce bandwidth usage and propagation delay, ensuring that even with many proposers, data swiftly reaches every validator.
• Implementation Strategies for Throughput and Latency: Practical engineering tactics that increase throughput and lower latency in a multi-proposer setting. This includes pipelining consensus rounds, batch processing of signatures/cryptography, and exploiting parallelism on modern hardware. Discussion of how small tweaks – like efficient handling of signatures, batching votes, or pruning DAG tails – can yield big wins in performance. Real-world examples (from Sui’s codebase and others) illustrate how these optimizations are applied.

• MEV Protection in Multi-Proposer Models: Exploration of how having many proposers can enhance fairness and reduce MEV (Maximal Extractable Value). Examination of how parallel block production limits any single validator’s ability to control transaction ordering, and how secret sharing or threshold encryption based methods can offer privacy. Attendees learned how consensus design can complement application-layer MEV defenses to create a more level and transparent playing field for transactions.

• Conference Website
• Recording
• Slides