MegaETH: An Upcoming High-Performance Layer 2 Project

 

MegaETH is an upcoming Layer 2 (L2) project that has been dubbed “Real-Time Ethereum” with sub-millisecond latency and the ability to process over 100,000 transactions per second (TPS). The project just announced $20 million in seed funding at a valuation of over $100 million!

 

Well-known Investors Support MegaETH

 

The financing was led by Dragonfly Capital, with participation from Ethereum founder Vitalik Buterin, Consensys founder Joe Lubin, Lido/Flashbots strategy lead Hasu, cryptocurrency trader Cobie, and EigenLayer founder Sreeram Kannan.

 

Innovation of MegaETH

 

MegaETH delivers industry-leading performance and decentralization guarantees by innovating on the existing Ethereum Virtual Machine (EVM) blockchain. MegaETH leverages Ethereum’s L2 technology to create differentiated roles for nodes with different hardware requirements, including sorters, provers, and full nodes.

High-performance implementation

MegaETH’s sorter is responsible for sorting and executing user transactions, passing state differences to full nodes through a peer-to-peer network. Full nodes apply these differences to update their local state instead of re-executing transactions to verify block integrity.

Systematic optimizations

MegaETH not only optimizes several components of the stack, but also identifies and solves numerous problems of existing blockchains, building a new system designed to approach the theoretical performance ceiling of decentralized blockchains.

In-memory computing and just-in-time compilation

MegaETH sorter stores its entire state in memory, enabling in-memory computing and improving state access speed. The just-in-time (JIT) compiler converts smart contract code into "native machine code", greatly improving execution speed and efficiency.

New state trie and efficient peer-to-peer protocol

MegaETH creates a new state trie that minimizes disk I/O operations and stores terabytes of state data. The efficient peer-to-peer protocol ensures that full nodes stay synchronized at the maximum update rate.