Layer Two and Rollups

note

This comparison covers general information regarding two widely used rollup mechanisms that are used to scale (usually EVM-based) blockchains and compares and contrasts how Polkadot achieves scalability.

Layer two (L2) networks are popular as being the way forward for blockchain scalability by off-loading the majority of computation from layer one (L1) networks. L2 solutions utilize the L1 network's security and functionality to build an additional layer that is often faster, reduces fees, and solves other platform-specific issues. In many cases, L2 solutions focus on utilizing block space on a particular blockchain efficiently and cost-effectively.

Rollups are an L2 scaling solution. At the most basic level, a rollup L2 solution is responsible for "rolling up" transactions by batching them before publishing them to the L1 chain, usually through a network of sequencers. This mechanism could include thousands of transactions in a single rollup.

Polkadot implements this functionality at the native level (i.e. without using L2 scaling solutions), allowing for shared security and scalability of the relay chain and respective parachains. Shared security is a concept that has similar goals to EVM-based optimistic and zero-knowledge rollups. Still, instead of being implemented as a secondary layer, Polkadot guarantees native security and scalability for each of its parachains through the Parachains Protocol. Polkadot handles the coordination of data from parachains into an aggregated, representative state, somewhat similar to L2 rollups.

Optimistic Rollups

Optimistic rollups are an interactive scaling method for L1 blockchains. They assume optimistically that every proposed transaction is valid by default.

In the case of mitigating potentially invalid transactions, optimistic rollups introduce a challenge period during which participants may challenge a suspect rollup. A fraud-proving scheme is in place to allow for several fraud proofs to be submitted. Those proofs could make the rollup valid or invalid. During the challenge period, state changes may be disputed, resolved, or included if no challenge is presented (and the required proofs are in place).

While optimistic rollups provide scalability, they have both benefits and drawbacks to their approach:

Benefits:

• They are not limited by the type of state change - any state change can be included, meaning existing apps do not have to account for it.
• They can be parallelized for scalability.
• A substantial amount of data can fit within a single rollup (in the case of Ethereum, for example, tens of thousands of transactions in a single state transition).

Drawbacks:

• Transaction censorship and centralization are of concern, where sequencers/L2 nodes can be compromised.
• Challenge periods could take a substantial amount of time to pass, increasing time for the rollup to finalize onto the L1 network.
• Due to their generalist nature of including any state change for their parent network, optimistic rollups can run into gas limitations or cause network congestion in the case of Ethereum.

Optimistic rollups are often used in the Ethereum ecosystem. Examples of optimistic EVM-based rollup solutions include:

• Optimism
• Arbitrum
• Unipig

Zero-knowledge Rollups

Zero-knowledge rollups (often called ZK rollups) are a non-interactive method that utilizes zero-knowledge proofs to compute the validity of a particular set of state changes. Whereas optimistic rollups relied on fraud proofs, ZK rollups rely on cryptographic validation in the form of ZK proofs.

Zero-knowledge rollups are significantly faster in finalization, as the cryptographic validity proof handles the nuance of ensuring a rollup is valid. However, the ZK rollups often suffer from performance due to their complexity and difficult implementation into resource-constrained environments. Because Turing completeness is also challenging to achieve due to this computational overhead, their ability to be generalized (in terms of blockspace) is reduced. However, they have a promising future in solving some of the problems of optimistic rollups and addressing secure scalability.

Benefits:

• They only require a small amount of data availability. Often, the proof is enough to ensure validity.
• They can be proven trustlessly.
• Because the proof is immediately available, finality is also instantaneous.
• They have a promising future overall, as they have not reached maturity yet.

Drawbacks:

• They suffer from the same problems that other L2 solutions have regarding the centralization of L2 operators.
• They are computationally expensive, and ZK circuits are difficult to implement.
• The potential for congestion is still a factor, as the amount of data could still be problematic.

Polkadot - Native Shared Security

Whereas rollups are considered solutions for L2 protocols, Polkadot include this functionality natively through its Parachains Protocol. The Parachains Protocol, which is how Polkadot handles network's sharding is meant to accomplish the combined goals of providing security, scalability, and availability.

It enables parachains to verify their collective state and communicate with one another. Parachains have similarities to aspects of optimistic and ZK rollups, which are reflected in how Polkadot handles the validity and availability of the parachain state. Collators, a key part of Polkadot architecture, are in principle similar to sequencers, as collators pass data with a proof-of-validity (PoV) function for liveness and communication with the relay chain.

Each shard, or parachain, is equipped with a unique state transition function (STF). This function ensures that communication to the relay chain remains valid. Each STF, called runtime, is written in Wasm. Any state transition function is valid if it compiles to Wasm and abides by the Parachains Protocol.

Each STF runs a validity proof. The proof (the Approval Protocol) is interactive, unlike ZK rollups, which are non-interactive. Additionally, unlike ZK rollups, there are no difficulties in creating parachains with Turing-complete logic. Each parachain is also a full-fledged state machine (usually in the form of a blockchain). Similarly to optimistic rollups, the Parachain Protocol also has cases where disputes and resolutions of potentially harmful para blocks (blocks representing the parachain) can take place, in which case the validators that vouched for that parablock are slashed if it is found to be bad.

Benefits:

• Protocol level sharding, shared security, and interoperability.
• Each shard has a low barrier of entry in terms of development, as anything that compiles to Wasm is a valid target.
• Fast Finality (usually under a minute on Polkadot).
• Data availability is built-in through validators and mechanisms like erasure coding.
• No L2 implies less of a risk of incurring centralization issues for sequencers or other L2 operators.

Drawbacks:

• Execution of code in Wasm could be a performance bottleneck, as it is slower than making native calls.
• The relay chain sets a hard limit on the size and weights of the PoV (Proof of Validity) blocks which contain the parachain state transition data.

Despite these drawbacks, Polkadot remains upgradable through forkless upgrades, which allows the protocol to be easily upgradable to stay in line with future technological advances.