Polygon zkEVM bytecode-level vs language-level zkEVM approaches
Quote90 words
Distinguishes language-level zkEVMs that transpile Solidity (Starknet, zkSync) from bytecode-level ones with no transpiler (Polygon zkEVM, Scroll), noting Scroll is closer to full EVM equivalence.
Key claims
Starknet and zkSync are language-level
Polygon zkEVM and Scroll are bytecode-level
Scroll is closer to EVM equivalence than Polygon
out of the two main #zkEVM types, @0xPolygonZK went for the ambitious bytecode level compatibility
- language level: devs can write code in Solidity but it gets transpiled into zk optimized language
- bytecode level: no transpiler step before code gets compiled & interpreted
so from the zkEVMs in the above image, #Starknet and @zksync would be considered language level compatIble
another #zk rollup taking the bytecode level approach is @Scroll_ZKP, which comes even closer to #EVM equivalence than #Polygon
for more on #zkEVMs, check my thread below ✨👇
x.com/expctchaos/status/1569717615282782209
QUOTING @0xPolygonFdn
Polygon #zkEVM was built around bytecode-level compatibility with the EVM, rather than language-level compatibility. But plenty of other ZK rollups take the language-level approach. So why was this design choice made?
Polygon #zkEVM was built around bytecode-level compatibility with the EVM, rather than language-level compatibility. But plenty of other ZK rollups take the language-level approach. So why was this design choice made?
1 image on X
What came next · since 11 Jan 2023
Since then
Zero has written 5 more pieces on Polygon zkEVM since this one.
- 2023 narratives: zkEVMs, privacy L1s, restaking, modular DA and perpsOn Polygon zkEVM
- Polygon zkEVM vs zkSync: organic versus airdrop-driven growthOn Polygon zkEVM
- Polygon zkEVM ecosystem TVL vs zkSync: the airdrop effectOn Polygon zkEVM
- Polygon zkEVM two weeks after launch: users, TPS and TVLOn Polygon zkEVM