As progress in the decentralized applications (dApps) ecosystem accelerates, scalability issues have become a critical bottleneck for Web3 progress. Here’s how modern Layer 2 scalability solutions for Ethereum attempt to improve the performance of the largest smart contract platform.

What is Scalability?

In terms of infrastructure, blockchains are decentralized systems of nodes designed to confirm transactions between accounts. The transaction is confirmed once the nodes agree to include it in the block.

Due to specifications of hardware technical capabilities, specific blockchain consensus characteristics, and the type of node software used, transactional throughput (usually measured by the number of transactions a blockchain can process each second, or TPS) of any blockchain has its limits.

At the same time, the pressure on the blockchain increases over time as new dApps are deployed on the chains and new crypto holders start using them for transactions. Thus, the speed of blockchain operations decreases while transaction fees increase.


For example, Bitcoin (BTC) can process three transactions per second while Ethereum (ETH) handles a maximum of 15 transactions per second.

In this case, the scalability of distributed networks should be viewed as their ability to handle increasing transaction loads, as well as increasing the number of nodes.

What are Layer 2 scalability solutions and why do dApps need them?

Ethereum inventor Vitalik Buterin introduced the so-called “Bitcoin trilemma” concept: he suggested that no blockchain can be scalable, decentralized and resource efficient at the same time. As such, truly decentralized and profitable blockchains are not the best in terms of scalability.

Image by Researchgate

During the acceleration of blockchain adoption in 2016-2017, engineers began discussing the possibility of introducing some systems that could work “on top” of existing blockchains to help them scale. Basically, they are designed to process certain data outside of the main blockchain.

They are connected to underlying blockchains (“layer 1” or L1 platforms) and partially verify the data through their instruments. As such, the blockchain public has referred to them as “layer 2” solutions or “second layer” systems.

Decentralized applications and end users need it to process blockchain transfers faster and pay lower transaction fees.

First generation of L2 solutions: Lightning from Bitcoin and Raiden from Ethereum

Pioneering Layer 2 scalability solutions used so-called “payment channels”: their elements were connected together to partially process data without sending it to L1s.

In 2015-2016, when some experts proposed to evolve Bitcoin (BTC) by increasing the size of its blocks, Professor Thaddeus Dryja invented the Lightning Network (LN). It is an L2 protocol based on payment channels. LN nodes transact with each other; periodically, the results of their operations are verified by the Bitcoin mainnet.

Bitcoin’s Lightning Network is still relevant in 2022; the aggregate volume of its channels is approaching 3,700 Bitcoin (BTC), as covered by U.Today.

Raiden network inspired by Bitcoin and Lightning, the first-ever channel-based solution for Ethereum (ETH), has been successful. His team raised $27 million via RDN ICO in 2017 and created a pioneering network scalability solution with smart contracts.

However, channel-based solutions also lack scalability; their capacity is limited by the number of nodes. As such, Ethereum (ETH) engineers introduced Plasma, a protocol that aggregates data from multiple transactions into a single dataset and processes it in Ethereum (ETH). The mainnet treats it as a single transaction; therefore, its TPS may be increased by thousands of percent.

Image from the Ethereum Foundation

Polygon (then Matic Network) was the first consumer implementation of Plasma. Due to the unforeseen interest in its solutions, in the first quarter of 2021, Polygon developers decided to rebrand it as a general-purpose L2 hub for DeFi.

Industry-Leading Layer 2 Scalability Protocols: U.Today’s 101

The 2021 bull rally triggered a new wave of crypto adoption: pressure on the L1s – largely on Ethereum (ETH) – increased and led to spikes in transaction fees. However, the main “Ethereum killers”, including BNB Chain, Solana, Fantom and Terra, are still criticized for their centralization problems. This is why Ethereum L2s are of particular importance for the sustainable progress of the DeFi ecosystem.


Mainnet launch: 2021

Technology: Optimistic aggregates

Bandwidth: 500-2,000 GST

Fees for token exchange: $0.49

Optimism is an Ethereum Layer 2 scalability network that leverages optimistic rollups. This technology “bundles” transactions and sends them to Ethereum (ETH) in compressed form. In Optimistic Ethereum, transactions are assumed to be valid by default. The validity of a transaction is assessed for fraud only in the event of a dispute.


Mainnet launch: 2021

Technology: Optimistic aggregates

Bandwidth: up to 4,500 TPS

Fees for token exchange: $0.91

Much like Optimism, Arbitrum uses optimistic rollups to help Ethereum (ETH) scale. However, it exploits different proof mechanisms. Instead of one-round cheat proof, Arbitrum uses multi-round cheat proofs: this reduces reliance on Ethereum (ETH), but such operations are charged with high fees. Arbitrum uses its own AVM (Arbitrum Virtual Machine), perfectly compatible with Ethereum Virtual Machine, or EVM.


Mainnet launch: 2021 (first embedded dApps in 2022)

Technology: Unconscious rollups, or ZK rollups

Bandwidth: 3,000 GST

Fees for token exchange: $0.09 (estimated)

StarkNet and its product StarkWare Alpha use a different technology: its rollups (validity rollups or ZK-rollups) are verified before being released on the Ethereum (ETH) mainnet. Its patented technology, ZK-STARK, moves all calculations to a single off-chain STARK prover, then verifies the integrity of those calculations using an on-chain STARK verifier.


Mainnet launch: 2020

Technology: Unconscious rollups, or ZK rollups

Bandwidth: 2,000 GST

Fees for token exchange: $0.24 (estimated)

zkSync is a general-purpose scalability protocol that allows every dApp to use the SNARK proof mechanism. It streams transactions to Ethereum’s L1 every 10 minutes and works natively with all non-custodial wallets for Ethereum and EVM compatible blockchains.


Mainnet launch: 2021

Technology: Optimistic rollups with improved solutions

Bandwidth: 1,000+ GST

Fees for token exchange: $0.14 (estimated)

Metis is a second-layer scalability solution for Ethereum (ETH) that prioritizes no-code experience for dApps developers. It is designed to evolve into a “plug-and-play” module for the next generation of DeFis, NFT marketplaces and decentralized autonomous companies (DACs); native crypto asset METIS is poised to support its tokenomics.

At the end of the line

Second-layer solutions should be viewed as a class of techniques designed to overcome the scalability limitations of mainstream blockchains.

Bitcoin’s Lightning Network and Ethereum’s Raiden pioneered L2 platforms.

Modern Ethereum-based Layer 2 protocols enable end users and developers to reduce transaction fees and increase bandwidth by thousands of percent. Optimistic rollups and ZK rollups are the two most common classes of L2 scaling instruments.