Spicenet Docs
  • Welcome to Spicenet
  • Introduction to Spicenet
    • Understanding the Need for a Global Financial Internet
    • The Sovereign Rollup Approach
  • Core Technical Architecture
    • The Nearly Optimal Merkle Tree (NOMT)
    • Asynchronous Execution Pipeline
    • Validator System Architecture
    • Shockwave
  • Advanced Technical Components
    • Sharding Implementation
    • Decentralized Sequencing
    • Based Multiple Concurrent Proposers
    • Multi-Staking Protocol
  • spicenet trading stack
    • Spicenet Trading Stack: Enshrined Liquidity and Market Structure
    • Unified Margin Accounts (UMAs)
    • Global and Local Markets
    • Instrument and Product Specification
    • Flashbooks: Intention-Based Orderbooks
    • Network-Owned Liquidity(NoL): A Deep Dive
    • Decoupled Order Matching and Settlement
  • spicenet execution network
    • Spicenet Execution Network
  • Products
    • PepperDEX: The Connective Tissue of Spicenet
    • hotUSD: the first Bitcoin dollar
    • User Abstraction Suite
    • Capsule: Smart Wallet Module
  • $SPICE Token
    • What is $SPICE?
    • $SPICE Utility
    • $SPICE Distribution
    • Airdrop and Ambassador Program
Powered by GitBook
On this page
  • Prelude
  • Multiple Concurrent Proposers Overview
  • Based MCP Implementation
  1. Advanced Technical Components

Based Multiple Concurrent Proposers

PreviousDecentralized SequencingNextMulti-Staking Protocol

Last updated 5 months ago

Prelude

Multiple Concurrent Proposers (MCP) represent a sophisticated design framework for blockchains that enables simultaneous block proposal by multiple leaders. Notable implementations include:

  • Sui

  • Aptos

These blockchains utilise MCP through Narwhal, their DAG-based block production protocol. This documentation outlines a Multiple Concurrent Proposer protocol specifically designed for Rollups and Layer 2 solutions, leveraging a Sovereign Rollup with Celestia as the underlying base layer for Data Availability (DA) and consensus guarantees.


Multiple Concurrent Proposers Overview

Core Concepts

  1. Simultaneous Leadership

    • Multiple leaders can propose blocks concurrently

    • Nodes receive and execute blocks incrementally from multiple leaders per slot

    • Eliminates traditional "leader monopoly" issues

  2. Block Production Mechanism

  • Leaders produce "local blocks" containing:

    • Transactions

    • References to previously produced blocks

  • Validators distribute local blocks network-wide

  • Universal block copy maintenance across validators

  1. Reconciliation Process

  • Key properties:

    • Deterministic

    • Universal

  • Ensures consistent transaction ordering across validators

  • Every validator decouples all newly received local batches into a single sequence of transactions, and re-orders them based on priority fees paid by the transactions. This ensures that every validator achieves a consistent transaction ordering outcome, and hence consistent execution results.

  1. State Synchronisation between Proposers

Each proposer propagates it's local batches to the network, including the other proposer. For example, the native proposer would propagate it's local batches to the based proposer, like it would to any other validator.

  1. This allows proposers to stay synchronised with each other while continually proposing divergent batches.

  2. Each proposer ensures that he is building the current batch on top of the latest synchronised state root.

Key Implications

MEV (Maximal Extractable Value)

  • Reduced validator control over MEV extraction

  • MEV extraction depends on final transaction ordering

  • Local block transaction ordering has minimal impact

Censorship Resistance

  • Elimination of leader monopoly

  • Enhanced transaction inclusion guarantees

  • Users can submit to multiple proposers

  • Duplicate transaction removal during reconciliation


Based MCP Implementation

Architecture

Based MCP introduces a hybrid approach combining:

  1. Based sequencing

  2. Native sequencing

Sequencing Types

Native Sequencing

  • Implementation: Sequencer run by Spicenet protocol

  • Current Status: Centralized

  • Future Development: Planned decentralization via auction mechanism

  • Characteristics:

    • Lower security guarantees

    • Lower liveness guarantees

    • High performance capabilities

Based Sequencing

  • Implementation: Permissionless L1 block proposer appointment

  • Function: Includes rollup batches in next L1 block

  • Characteristics:

    • Higher security (L1 stake-backed)

    • Enhanced liveness guarantees

    • Longer inclusion and pre-confirmation times

Performance Trade-offs

  • Native sequencer offers superior performance metrics

  • L1 sequencer requires additional processing time for:

    • Inclusion

    • Pre-confirmation

    • Rollup block integration into L1 blocks