Thread by @0x_Specter

Saw THE NERDIEST THING EVER --> Friend downed a cup of gin over a ZK-EVM debate at 2am the other day.

Figured I better do my homework before meeting them next.

Mini 🧵summarizing @VitalikButerin 's "The Different Types of ZK-EVMs"
vitalik.ca/general/2022/0…

@VitalikButerin 1/ There's been a lot of buzz recently about "ZK-EVM" projects.

Projects like @0xPolygon, @zksync, and @Scroll_ZKP have all made announcements, alongside others.

We will explore key differentiation between ZK-EVM & hopefully it brings your zero-knowledge up to some-knowledge!

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 2/ The goal of these projects is to use ZK-SNARK technology to make cryptographic proofs of execution of Ethereum-like transactions.

This could make it much easier to verify the Ethereum chain itself or to build ZK-rollups that are much more scalable.

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 3/ But there are subtle differences between these projects, and what tradeoffs they are making between practicality and speed.

In this 🧵, we'll describe a taxonomy of different "types" of EVM equivalence, and what are the benefits and costs of trying to achieve each type.

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 4/ Ethereum equivalence refers to degree of similarity between L1 & L2 solutions.

EVM equivalence describes the compliance of L2 solutions with the Ethereum yellow paper.

The goal of EVM equivalence is to ensure that any bytecode that can be ran on L1 can also be ran on L2

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 5/ Type 1 (Fully Ethereum-equivalent)

Type 1 do not change any part of the Ethereum ecosystem and strives for perfect compatibility to verify blocks as they are

Bootstrapped by community contributors including @Scroll_ZKP, Taiko & others

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 6/ Type 1s are seen as the most ideal way to scale Ethereum as is. Additionally, current Ethereum infrastructure is fully reusable on Type 1s

However, Ethereum was not originally designed around ZK-friendliness, so producing proofs for blocks currently takes many hours.

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 6/ Type 2 (Fully Ethereum-equivalent)

Type 2 means some minor modifications to Ethereum to make development easier and proof generation faster.

@Scroll_ZKP- project is building toward a Type 2, as is
@0xPolygon (Hermez)

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 7/ Type 2s provide faster prover times than Type 1s by removing parts of the Ethereum stack that rely on needlessly complicated and ZK-unfriendly cryptography.

However, they are still slower than we would like.

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 8/ Type 3 (almost EVM-equivalent)

Type 3 means significant modifications to the Ethereum system to make ZK-friendly.

Type 3s are the fastest but has the most tradeoffs.

Scroll & Polygon are both Type 3 in current forms, interesting to see compatibility improvements to come

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP 9/ TLDR

Type 1 ZK-EVMs have perfect compatibility but slow prover times.

Type 2 ZK-EVMs provide faster prover times but still have some compatibility issues.

Type 3 ZK-EVMs are the fastest but make more significant modifications to Ethereum.

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP VIRGIN THREAD but I hope you've found this thread helpful. Let me know your thoughts and comments.

Follow me @TheDegenBanker_ for summaries of readings I come across while plzfixing in the day & thoughts on Web3 M&A and fund raising

Like/Retweet the first tweet below PLEASE:

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP tagging some senpais because i love to plzfix:
@Defi_Maestro
@LandfSmile

@cornergirl999
@0xavarek
@chiayong_
@0xPoops

@0xjlow (sigma male scammer)

@VitalikButerin @0xPolygon @zksync @Scroll_ZKP @Defi_Maestro @LandfSmile @cornergirl999 @0xavarek @chiayong_ @0xPoops @0xjlow #ZK-EVM #0xPolygon #zksync #Scroll_ZKP #VitalikButerin #Ethereum #L1 #L2 #Type1 #Type2 #Type3