How the Merge opens up Ethereum to sharding and massive scalability

There’s a common misconception that changes implemented on the Ethereum blockchain in Q4 2022 will result in low gas fees. But only the introduction of sharding will reduce transaction fees natively on the Ethereum blockchain.

So what is sharding and when can we expect it to be introduced?

What is sharding?

The sharding of a blockchain involves splitting the blockchain up into smaller and more manageable segments. By storing data in independent, separate shards, the computational burden on the overall network is reduced. Collectively, all shards in a blockchain network form one logical dataset.

Sharding is not a new concept. It has been utilized within conventional database design for some time. ‘Shard’ is an abbreviation for ‘System for Highly Available Replicated Data’ - the name of an 1980s database product.

Through sharding, a blockchain network like that of Ethereum can improve network latency (the speed at which a bit of data moves on the network) and scalability.

In its current form, Ethereum achieves around 13 transactions per second (TPS). We won’t know for sure until after sharding has been implemented but it’s thought that throughput of 100,000 TPS may be possible on the network.

How does sharding work?

To understand sharding, we first need to consider how nodes work.

Nodes play a fundamental part in a blockchain network. Without them, a blockchain’s data would be inaccessible. They have three main functions: validating transactions, communicating that validation around the network, and storing the history of the network’s transactions.

In the case of proof-of-work based blockchains like Bitcoin, sharding is incredibly difficult to achieve as nodes need information from the entire network.

Ethereum’s move to a proof-of-stake consensus mechanism means that the nodes no longer have to store the entirety of the network’s activities. Each node only has to maintain data related to its own shard. Information still has to be shared from a shard to other network nodes though in order to maintain network security.

Phase 2 of Ethereum 2.0 will involve a physically sharded system of 64 linked databases. Transactions are processed simultaneously and in parallel with each shard. Network nodes will only process certain specific operations and not all operations like with the current model.

Notaries are picked at random within each shard to vote on the validity of a shard block i.e. a batch of transactions. Each shard will have a ‘committee’ of 128 validators. The committee proposes and validates each block every 12 seconds via a shard manager contract.

Each individual shard chain will have a set of nodes assigned to it to validate new transactions. With 64 shards enabling simultaneous parallel processing, significant performance improvements can be achieved in terms of throughput and latency.

With improved network efficiency and performance, the network will scale to handle increased user demand.

Timeline

Sharding is not scheduled to be implemented on Ethereum until H2, 2023 as part of ‘The Surge’, phase 2 of the Ethereum 2.0 project.

We should be guarded in our expectations when it comes to that timeline given past delays. The Merge  - the first phase of Ethereum 2.0 which will see the network move from Proof-of-Work to a Proof-of-Stake consensus mechanism - has been delayed six times since 2017.

Lack of scalability has been a major obstacle in the development of Ethereum. High gas fees have led to a whole host of alternative smart contract blockchains emerging as potential ‘Ethereum killers’.

Those chains have garnered some momentum but comparatively, Ethereum still has by far and away the greatest concentration of developer activity.

Interim solutions

In April 2022, Tim Beiko, the Ethereum Foundation Coordinator for Core Developers, said that there is a possibility that developers will choose to implement an interim solution.

A new type of Ethereum transaction that has a small proportion of shard data on it could be implemented as set out in Ethereum Improvement Proposal (EIP) 4844. This is being considered and may be introduced in a way that allows the subsequent roll-out of the complete sharding infrastructure at a later stage.Meanwhile, the ongoing development of layer 2 off-chain scaling solutions such as ZK-rollups and optimistic rollups is interesting and they may act to some degree as a stopgap solution in dampening down gas fees on Ethereum.

In reality though, those solutions need sharding in order to enable cheaper and bigger rollups. Once both are in place, Ethereum should be set for a Cambrian explosion of dApps, Web3 tooling and mass adoption.