- Monolithic blockchains are managing 3 layers: Consensus/security, execution, and data availability. This is hard. And it doesn’t scale well or easily. A modular design that compartmentalizes and optimizes these functions will work better.
- Solana and Avalanche aren’t competing with Ethereum — they’re competing with ZkSync, Arbitrum and StarkNet.
- We will have security and decentralization that’s practically identical to Ethereum, with massive scalability.
- We can design & iterate on our own governance structures that last forever, are 1000x as efficient to operate, and cost 1000x less to deploy.
Things To Ape (if they launch tokens)
Composable smart contracts on zkSync2. Will also introduce zkPorter, a PoS chain that allows zkSync2 users to publish data on zkPorter instead of Ethereum, enabling higher transaction throughput with slightly diminished security guarantees. Recently added functionality for NFTs, including minting, swapping, transferring, and withdrawing to Ethereum.
Arbitrum is a suite of Ethereum scaling solutions that enables high-throughput, low-cost smart contracts while remaining trustlessly secure.
Optimistic rollups that assume submitted transactions are valid and rely on a 7 day fraud proof for users to contest potentially fraudulent transactions.
Optimistic-inspired rollup. They’ve reduced the fraud-proof window to only a few minutes.
Pluggable consensus and data availability layer. Quickly deploy a decentralized blockchain without the overhead of bootstrapping a new consensus network.
General-Purpose, scalable data-availability layer.
Current projects actively pursuing a modular framework:
Blockchain’s three broad functions: consensus (or security), data availability, and execution can be split arbitrarily. This will speed things up, a lot.
The new paradigm of modular blockchain legos is upon us.
Soon, rollups will allow users to perform complex DeFi transactions 90%-99% cheaper than on Ethereum’s mainnet.
Data shards will also lower costs significantly, but we may need to wait another year or two.
Execution layer — ZkSync, Starknet, Arbitrum
This is where end users will interact with various protocols to perform swaps and other smart contract transactions.
Consensus layer — Ethereum’s Beacon chain, Polkadot’s Relay chain
This is how the network maintains it’s security. Bundled rollup transactions are submitted to a larger, more secure L1 network.
Data availability layer — Polygon-Avail, Celestia
This layer is the history book and the source of censorship-resistant truth, guaranteeing that all transactions occurring on rollups can be proven and referenced.
There are now L1s with no execution layers. Their purpose is just to be a data availability layer for other chains, that’s it. They have their own consensus mechanisms just for data, and they’re useless by themselves.
Rollups are how modular chains like Ethereum will scale. They are blockchains that are only focused on executing transactions as fast as possible and outsource the work of security and data availability to a layer 1 chain that’s optimized for it.
It works by bundling transactions together, compressing them, and sending them to a consensus layer to be verified. This increases efficiency by verifying multiple transactions at once while also scaling by increasing the number of transactions that can be executed. It remains fully composable, even if it’s settled across multiple data shards or external data availability sources.
“Rollups are simply constructions that can get the best of all worlds — with high scalability, security, and decentralization.” — Polynya discussing common rollup misconceptions.
The current tech is roughly split into 4 categories:
- Snarks or starks that keep data on chain: ZK rollup
- Fraud proofs that keep data on chain: Optimistic rollup
- Snarks or starks that keep data off chain: Validium
- Fraud proofs that keep data off chain: Plasma
Early rollups came with “training wheels” which made them trusted and permissioned in certain respects. Over time, these will be replaced with fully decentralized, trustless and permissionless platforms.
The final form of rollups will likely be Zkrollups with decentralized sequencers, decentralized provers, decentralized L1 smart contracts and light unassisted exits. More from Polynya on this topic in Volitions — best of all worlds and A new paradigm of blockchain construction.
Optimistic Rollups Explained
Optimistic rollups just assume that transactions are valid. They let the operators of the rollup post any data they want, and rely on a 7-day time window for potentially fraudulent transactions to be discovered, challenged, and reverted.
Zero Knowledge Rollups Explained
While Optimistic rollups assume transactions are valid and leave room for others to prove fraud, ZK rollups actually prove the transactions are valid. They also have the ability to integrate privacy into the rollup, and there’s no need for 7-day fraud proof windows or LP incentivized quick-withdraw bridges.
Succinct Non-interactive Arguments of Knowledge.
- Small and efficient
- Fast to create and verify proofs
- The smaller proof size also requires less gas to verify when submitted to L1
- Assumed to be vulnerable to quantum attacks
zkSNARKs require trust assumptions as they need to agree upon a set of parameters for a trusted setup, but new systems have emerged that don’t require an initial trusted setup.
Scalable Transparent Arguments of Knowledge.
- The size is slightly inhibitive, requiring more gas to submit the proof to L1
- They’re quicker to generate and will scale more efficiently than Snarks as computation requirements increase
- Quantum resistant
More info here.
What happens to existing L1s?
We’ll probably just keep optimizing lots of existing and new chains for each of these 3 specializations. Existing monolithic chains will start to divide their responsibilities. Some will become data availability layers and others will become rollups. As investors, we should be asking ourselves which existing networks have a current head start and willingness to focus on specialization, and be watching new teams that are advancing this technology.