Monad: The Most Anticipated Parallel Blockchain in 2024

However, despite Ethereum's prominence, it and its associated blockchains suffer from low transaction speeds: Ethereum at 13 TPS, Arbitrum at 12 TPS, and Optimism at 6 TPS.

MonadBFT serves as the consensus mechanism for Monad, ensuring transaction order integrity. This mechanism embodies Byzantine Fault Tolerance (BFT), allowing the decentralized system to function accurately even amid component failures or malicious actions.

For a clearer understanding of MonadBFT's advantages, it's helpful to contrast it with Cosmos' Tendermint consensus protocol. Tendermint's block finalization demands three distinct phases: 

• Proposal;
• Pre-vote;
• Pre-commit. 

This entails a leader proposing a block and disseminating it among validators for two rounds of voting. A consensus among validators is needed for the block to be officially recorded on the blockchain. This methodology introduces extra time requirements, as nodes must finalize the current block before moving on to the next.

Conversely, MonadBFT adopts a two-phase process utilizing a fan-out-fan-in pattern, enabling parallel function execution. Block finalization within Monad can be delineated into:

• Commit Phase 1: The leader proposes block k and circulates it to validators. Each validator forwards their vote to the leader of the subsequent block, k+1. This leader compiles these votes into a Quorum Certificate (QC), evidencing the consensus of at least two-thirds of the validators.
• Commit Phase 2: The next block's leader, k+1, distributes the new block alongside the QC to validators, who then commit block k.

This process leverages a pipeline structure, blending the creation efforts for blocks k and k+1. Such a strategy diminishes the time needed to produce a single block by reducing the steps needed for communication.

In Ethereum and most other blockchains, transactions must be executed before the network can reach consensus. Monad adopts a different approach known as deferred execution, where Monad validators agree on the network state before executing transactions.

Simplifying the concept, deferred execution in Monad can be likened to the operations of a student cafeteria.

In most cafeterias, students first choose their food (transaction execution) and then line up to pay (consensus). If resources for preparing the food run out or if preparation takes too much time, the whole process can stall.

Monad’s “cafeteria” operates differently:

• Consensus: Before getting their food, students line up and tell the cashier what dish they will order. The cashier records the order of dishes, similar to how Monad records the order of transactions.
• Execution: After placing their orders, students proceed to the pickup counter. While one student is receiving their food, the cashier is already taking the next student's order.

This way, even if preparing a dish takes longer, the queue at the cashier continues to move. All students know what they will receive, and their order is already queued.

However, what happens if one of the nodes turns malicious (e.g., sets the network state randomly)? 

To address this issue, Monad utilizes a delay in confirming the Merkle root (the blockchain's final state value), delayed by D-blocks — a system parameter currently set at 10. This means the state of block N-D is only valid when the network reaches a consensus on block N.

A node with an execution error in block N-D will cease to participate in consensus starting from block N. This rolls back the blockchain state to block N-D-1, followed by re-executing transactions in blocks N-D, N-D+1, N-D+2, etc.

The challenge of parallel transaction execution lies in identifying dependencies between operations. Monad employs an optimistic execution method, where validators process transactions even before the processing of previous ones concludes.

This approach is sound when transactions are unrelated. However, scenarios can arise where the outcome might be incorrect, for example:

• Transaction 1 reads and updates account A’s balance (e.g., receiving a transfer from account B);
• Transaction 2 also reads and updates account A’s balance (e.g., transferring funds to account C).

With parallel execution, the processing of transaction 2 begins before transaction 1 is completed. Consequently, in sequential execution, the balance value used for account A could differ.

In the optimistic approach, the input data of transaction 2 is compared with the output data of transaction 1. If they differ, transaction 2 is re-executed with the correct data.

For a more detailed explanation of parallel execution in other blockchains, refer to this article.

Parallel execution in isolation doesn't enhance blockchain performance; what's needed is a database system capable of parallel read and write operations to the disk. 

Traditional blockchain data storage libraries, such as LevelDB or LMDB, lack support for executing transactions in parallel. This necessity led to the creation of Monad's proprietary database: MonadDB. Despite expanding blockchain capabilities, MonadDB maintains comparable read costs to the aforementioned LevelDB.

The project team behind Monad boasts a transaction processing capability of 10,000 TPS (transactions per second), making it a prime candidate for trading protocols, particularly decentralized exchanges. Beyond this, the blockchain finds applicability in GameFi, NFTs, oracles, and more.

• Wormhole: A multichain protocol facilitating asset exchanges across different blockchains, including Ethereum, Solana, and Avalanche, among others. 
• LayerZero: A protocol for cross-chain transfers, whose token is among the most anticipated assets for the years 2023-2024.
• iZUMi Finance: A multichain DeFi protocol with a daily trading volume of $13.6 million. 

The Monad ecosystem is continually expanding. Events like the Wormhole airdrop, which rewarded Monad community members with approximately 8,000 W (~$12,000) for their engagement on Discord, have positively impacted the project's visibility.

Monad is intriguing not just from a technological standpoint but also for its retroactive potential. A project of such magnitude could feasibly replicate the success seen by projects like Sei and Sui, in terms of token price growth, and Wormhole, in terms of airdrop success.

Follow our X account for the latest updates!