An In-depth Look at Ethereum’s Path to Scalability

After Ethereum was successfully launched in 2015, it came face to face with a couple of problems that restricted it from reaching its potential. The high demand led to a clogged network with high costs of transactions, rapid growth made the process of running a node hard and the Proof-of-Work consensus mechanism was not making Ethereum sustainable.

To combat these problems, Ethereum set out to upgrade its network and divided the upgrades into 5 phases namely; the Merge, Surge, Verge, Purge, Splurge

It should be noted that these phases are happening in parallel and not sequentially.

The Merge

The first phase in Ethereum’s grand plan at scaling happened on September 15, 2022, and this signaled the transition of Ethereum from the energy-hungry Proof-of-Work model to the green Proof-of-Stake model.

What happened when the merge occurred was that the mainnet which was the execution layer of Ethereum running on PoW was joined to the Beacon Chain which is the new PoS consensus layer. Now the network is secured by staking ETH.

The Beacon Chain

The Beacon Chain was shipped on December 1, 2020. It was the PoS blockchain that ran alongside the PoW Ethereum blockchain. It was created to test-run the Proof-of-Stake logic, ensuring that it was sound before it was integrated into the Mainnet. Now that the Merge has happened, there are no two blockchains but just one PoS Ethereum Chain.

Outcome Of The Merge

1. Increased Sustainability

The fundamental restructuring of Ethereum has dropped the energy requirements by 99.95% which now makes Ethereum more sustainable. This restructuring has now made it possible for Ethereum to be upgradable.

1. Improved Security

Proof of Stake has now democratized network participation and this now leads to a higher level of security. The Merge now makes it possible for anyone with an internet connection and enough ETH balance to participate in the network. Ethereum has more validators than all PoS blockchains combined and the amount of ETH staked makes Ethereum virtually too expensive to be attacked.

1. Reduction in ETH Supply

Under the PoW model, the amount of ETH released into circulation was about 13,000 ETH per day (mining rewards) but now that the Merge has happened, the amount of ETH released into circulation is estimated to be about 1,600 ETH per day (staking rewards). This drops the issuance of new ETH by approximately 90%. Also, Ethereum plans to burn about 1,600 ETH every day.

The Surge

The surge phase is the phase where Ethereum introduces sharding. The sharding concept is not specific to the blockchain world, it is actually a concept in computer science where a database is broken down into smaller parts known as shards in order to reduce the load on the main database.

The purpose of sharding is to increase transaction speed and to also reduce network congestion. In the case of Ethereum, when sharding is carried out, the network will split into 64 shards and each shard acts as a mini-blockchain containing its own state, account balances and smart contracts.

Validators will no longer need to verify every transaction being broadcasted on the Ethereum network, instead, they will focus on verifying new blocks on their own shards after which a committee of 128 validators then confirms on the main chain.

Features of Sharding

1. Everyone can run a node

With sharding in place, scaling can be achieved without trading decentralization because hardware requirements are reduced. Due to the fact that validators no longer need to store the entire data of the Ethereum database, the cost of storing data is reduced.

1. Increased network participation

Due to lower hardware requirements, one can run Ethereum using their personal computer or phone. This reduces the barrier of entry and increases security of the network since now, anyone can participate.

Shard Chains Version

Version 1: Data Availability

Initially, when the shard chains are launched, they won’t immediately be able to process transactions and handle smart contracts. Their responsibility will be to provide extra data to the network but still, when this version is combined with rollups, there is a tremendous improvement in transaction throughput.

Version 2: Code execution

In this version, extra functionalities are added to the shard chains to make them much more identical to the mainnet. They will be able to execute code, handle transactions, and contain their own unique set of account balances and smart contracts. There will also be cross-shard communication.

The need to upgrade shard chains from version 1 to 2 is being debated in the community considering the boost version 1 provides (i.e 100,000 transactions per second)

Initially, Ethereum thought of carrying out sharding before the merge but with the rise of L2 scaling solutions, it become clear that transitioning to PoS was more important.

Ethereum without rollups does between 15-40 tps, rollups on their own do 1,500-4,000 tps but when sharding is combined with the power of rollups, we can have about 100,000 tps.

The Verge

The Verge is the phase where verkle trees are introduced. Verkle trees are a powerful upgrade to Merkle trees. The Verge aims to optimize storage and also reduce node size. For context, we need to understand what Merkle trees are.

What are Merkle Trees?

Merkle trees can be described as binary tree that uses cryptographic hash algorithms. They efficiently verify the correctness of data and at the same time reduce the verification’s memory requirements. One major advantage of Merkle trees is that compared to other data structures, Merkle trees take up less space.

Ethereum currently uses a Merkle tree called Merkle Patricia tree which provides a data structure that can be used to store all key, and value bindings and is cryptographically authenticated.

What are Verkle Trees?

The Verkle trees are very much similar ro Merkle trees but their most important feature is their proof size efficiency. A Merkle tree will require about 1kilobyte to provide a proof for a tree with 1 billion data points but a Verkle tree will require less than 150 bytes. Relying upon polynomial functions, Verkle trees describe data using Polynomial Commitments.

As a result of its superior proof size, Verkle trees reduce the amount of complex cryptography required and also create an opportunity to make the Ethereum blockchain more efficient and scalable.

The Purge

The Purge is regarded as the cleanup phase where historical data is removed to reduce network congestion. Vitalik said that the purge is aimed at cutting down the amount of space you need to have on your hard drive, trying to simplify Ethereum, and not requiring nodes to store history.

As earlier mentioned, the reduction in hardware requirements will increase decentralization and in effect boost the security of the network.

The Splurge

This phase is regarded by Vitalik as “the fun stuff”. it aims at making the network run smoothly after all other upgrades have taken place.

Conclusion

In the future when Ethereum is 100% complete, it will look like an aggregation of L2 blockchains, data availability protocols, and side-chains. The future of Ethereum is bright, and I'm excited.

Also published here.