Monad successfully raised $224 million in 2024 — the largest fundraising round since the bull market of 2021.
With this massive capital, Monad is heavily focused on infrastructure, aiming to become one of the fastest and most efficient blockchains today. The network has achieved a peak throughput of 10,000 transactions per second (TPS), while peers like Sui, Aptos, and Solana typically reach between 2,000–6,000 TPS.
According to the development team, Monad’s technology introduces several significant innovations compared to other Layer 1 blockchains of the same generation, including:
- MonadBFT
- RaptorCast
- Asynchronous Execution
- Parallel Execution
- MonadDB
However, cutting-edge technology often comes with complexity. This article will help you better understand each of these components through a personified analogy — making them more accessible and easier to grasp.
How Monad Works
RaptorCast
RaptorCast is Monad’s custom data propagation protocol used to transmit information from the leading validator (leader) to other validators across the network.
When a transaction is received, the leader first applies Erasure Coding to break the data into multiple fragments. These fragments are then relayed through an intermediary entity, which helps distribute the data to the remaining validators in the network.
In traditional networks like Ethereum, the data propagation process is different. The data is typically sent from one validator to a nearby group of validators, which then further relay it to other neighboring groups. This approach resembles how a rumor spreads—starting within one neighborhood and gradually reaching the next.
Monad’s Co-Founder, Keone Hon, once likened RaptorCast to the process of moving a house. Instead of transporting the entire house from point A to point B, RaptorCast “disassembles” it and distributes the pieces across multiple delivery trucks. As long as at least two-thirds of the fragments arrive intact, RaptorCast can fully reconstruct the house in its original form.
With this model, RaptorCast helps Monad save bandwidth and accelerate transaction processing by enabling parallel data transmission to multiple recipients. It also enhances the scalability of the network.
MonadBFT
MonadBFT is Monad’s consensus mechanism, consisting of three main stages in the block formation process:
- Propose: The network randomly selects a validator to act as the leader. The leader breaks down the block data and distributes it to other validators using RaptorCast.
- Pre-vote: Validators vote to approve the received data. If at least two-thirds of the votes are in favor, the data is accepted.
- Finalize: The block is officially formed and permanently stored on the network.
This mechanism follows the Byzantine Fault Tolerance (BFT) model, which is widely used in networks like Aptos, Sui, and Cosmos. However, MonadBFT introduces three key improvements over traditional BFT.
First, Monad enables the entire process to run in parallel. This means that even if a block hasn’t completed the Finalize stage, the network can already select a new leader and begin processing the next block. This parallelization significantly boosts transaction throughput and input capacity.
Second, if the current leader becomes unresponsive during the process, the network automatically rotates leadership and assigns a new validator to take over — ensuring continuity and fault tolerance.
Monad Pipeline. Image: Monad
Asynchronous Execution
Asynchronous Execution is a mechanism that allows transaction execution and consensus to occur independently of each other. In simpler blockchain architectures like Ethereum or Cosmos, the transaction flow typically follows a linear process:
Transaction occurs → Consensus → Execution.
In normal conditions, this “single-lane” model has a critical drawback: each transaction must be fully completed before the next one can begin. This creates a bottleneck under high user load, slowing down transaction processing across the network.
With Asynchronous Execution, Monad enables a second transaction to proceed to consensus and execution without having to wait for the first one to finish. This is similar to setting up multiple ticket checkpoints at an event—improving processing efficiency and allowing users (transactions) to flow continuously without congestion.
MonadDB
MonadDB is a custom-built database developed by Monad, rather than relying on commonly used solutions like RocksDB or LevelDB. It was designed specifically to keep up with the high transaction throughput of the Monad blockchain.
Typically, when validators verify data or detect anomalies in a specific transaction, they need to query the database to cross-check the information.
As a result, any blockchain aiming to increase transaction processing speed must also have a database system fast enough to support the network’s continuous and rapid query demands — and MonadDB was built precisely for that purpose.
Cutting-edge technology — so why isn’t anyone adopting it?
Monad has achieved a peak throughput of 10,000 TPS, processing millions of transactions simultaneously — a figure many blockchains can only dream of. Yet, the question remains: why haven’t more projects adopted Monad’s innovations, and why is Monad the one pushing these boundaries?
First, achieving high TPS isn’t just about having fast software. It requires a fully coordinated infrastructure — from consensus mechanisms and data propagation to transaction execution and database architecture.
Monad is one of the very few projects that chose to rebuild everything from scratch. Rather than patching legacy systems, the team redesigned the entire chain to optimize for parallel performance.
Second, integrating new mechanisms like Asynchronous Execution, Erasure Coding, or a custom database like MonadDB comes with significant technical risks. These innovations require deep expertise in distributed systems — which is why few teams dare to pursue them, and why Monad stands out.
Finally, raising $224 million before even launching its testnet gave Monad the resources to build from the ground up and fine-tune every technical detail. For comparison, Solana raised under $50 million pre-testnet, while Avalanche raised around $90 million.
