Ethereum 2.0, Phase 1: Boosting Throughput with Sharding

New Tools to Boost Ethereum Blockchain Throughput

The Ethereum network is in need of increased throughput to keep up with its exponential rate of growth and platform development. To this end, Ethereum is currently undergoing a drastic and lengthy upgrade to what’s been dubbed Ethereum 2.0 — a redesigned Ethereum blockchain complete with a new Proof-of-Stake (PoS) consensus mechanism and the ability to process significantly more transactions per second (TPS). The Ethereum 2.0 updates intend to provide scalability that is orders of magnitude greater than what is currently available with Ethereum — a jump from roughly 10 TPS to potentially 100,000 TPS.

The entire Ethereum 2.0 update — called Serenity — has been divided into multiple phases due to its structural complexity, and to reduce the risk of technical issues that may arise from implementing too many changes at once. With this in mind, many phases of Serenity are being concurrently developed and will initially operate parallel to, but separate from, the legacy Ethereum blockchain until the final phases, which is when the new technical upgrades will eventually be integrated.

Phase 0, the first phase, was implemented in December 2020 and established the groundwork for the Beacon Chain — the new blockchain which will eventually manage Ethereum’s PoS consensus mechanism, validator nodes, and forthcoming shard chains. Phase 1 is next on the roadmap, and will include more tools designed to alleviate pressure on the Ethereum ecosystem. The key Phase 1 developments are shard chains, and a more recent development known as roll-ups.

Divide and Conquer: What Are Shard Chains?

The Ethereum 1.0 blockchain is one single consecutive chain of blocks that hold the transactional data of the network. It utilizes a Proof-of-Work (PoW) consensus mechanism where transactions are packaged into blocks, and each network node (all several thousand of them) must process and verify each new block. The system is very secure, but it can also cause transactions to become slow and cumbersome, providing an opportunity for it to benefit from more efficient transaction processing. Along with the transition to an Ethereum PoS network to boost transaction throughput, sharding has been proposed as an effective scaling solution.

Sharding is a form of data partitioning that breaks down a large database into smaller, more manageable pieces to boost the efficiency of the whole. In the case of a blockchain, shard chains are created by sharding, or partitioning, the blockchain’s data processing responsibilities into smaller pieces. This decreases the workload of network validators by requiring each validator to only store and manage one shard of the network instead of the whole blockchain. In doing so, shard chains enable parallel processing, reducing the latency of linear processing when using only a single blockchain.

Phase 1 of Ethereum’s Serenity upgrade will break down the single Ethereum blockchain into 64 shard chains, which will be coordinated by the Ethereum 2.0 Beacon Chain. Teams of validators — which have deposited collateral into the Ethereum 2.0 deposit contract as part of Phase 0 — will eventually be randomly assigned to manage particular shard chains on the Ethereum network.

Each shard chain will be assigned its own transaction group as well as unique tasks and responsibilities. In time, shard chains will be able to execute code — supporting smart contracts and decentralized applications (dApps), much like the main Ethereum blockchain — though at first they will merely test the aggregation and movement of data between shard chains and the Beacon Chain. It is also possible that even more shard chains will be created in the future as needed.

Challenges of Ethereum Sharding

Although the implementation of shard chains has the opportunity to greatly improve scalability, it also raises some concerns which will need to be adequately addressed before the upgrade can be truly operational. For example, because each shard chain functions as its own unique blockchain, internal communication and compatibility between them must be established. In this case, the exact relationship between shard chains and the Beacon Chain must be nailed down. Ethereum plans to implement its “receipt paradigm,” in which each shard chain will generate receipts of its respective internal transactions. Receipts will then be stored in the Beacon Chain’s shared memory — available for all other shards to view, but not tamper with.

While it’s true that establishing unique shard chains enables parallel processing on the Ethereum network, latency can arise when data needs to be shared between the shards themselves. For instance, if someone sends 50 ETH from shard A to a recipient on shard B, the process must be optimized to avoid sacrificing the increased throughput that sharding aims to provide — potentially slowing down transactions and nullifying some of the benefits of sharding altogether.

There are also potential concerns about increased security risks, as it would be much easier for a malicious actor to take over one shard chain as opposed to the base layer blockchain — a concern that Ethereum hopes to mitigate by requiring a much higher number of validators to support Ethereum 2.0 than its predecessor. With these concerns in mind, Phase 1 of Ethereum Serenity is due to be a trial run of sharding rather than a full-scale implementation.

Roll-ups in Addition to Shards: What Are Roll-ups?

Another strategy being developed as part of Serenity’s Phase 1 update is the implementation of roll-ups. Roll-ups, which already exist today, are a Layer-2 scaling solution that can operate in tandem with sharding.

As a Layer-2 strategy, roll-ups seek to move the brunt of data processing off of the base layer blockchain using extra architecture, rather than make changes to the base layer blockchain itself. The goal of a Layer-2 scaling solution is to only minimally interact with the base blockchain for finality and settlement, while enabling other smaller transactions to take place more efficiently off-chain.

Roll-ups are especially useful in boosting throughput of dApps because they make smart contract data processing more efficient — allowing dApps to aggregate, or “roll up”, transactions, bundling them into one single transaction off-chain before settling the final state on-chain. This is done to lower transaction fees since the rising popularity of Ethereum dApps and decentralized finance (DeFi) offerings have caused gas fees to increase substantially.

When Will Phase 1 Be Implemented?

Phase 1 of Ethereum Serenity will sequentially follow Phase 0 and is estimated to arrive sometime in 2021, likely in the latter half of the year barring any major development setbacks.

While different phases of the Serenity update package are being developed concurrently and will operate independently from the legacy Ethereum blockchain, some may be implemented sooner than others. The full integration of the Beacon Chain and its accompanying PoS consensus mechanism (Phase 0), and sharding architecture (Phase 1) likely will not go live until the final stages of Ethereum 2.0 development. Some, however, are hopeful that perhaps roll-ups can be implemented soon on Ethereum 1.0 and alleviate network congestion in the meantime.