What is Kakarot zkEVM? Why Vitalik participated in the vote

Author: s, venture capital analyst; translation: Jinse Finance xiaozou

What is Kakarot zkEVM? Why is it supported by Vitalik and StarkWare? Kakarot brings EVM to Starknet, facilitating the expansion of EVM dApps. Let's take a deeper look!

We've booted up and entered Super Saiyan mode. Kakarot Labs has been formally established and completed pre-seed early financing with top investors.

Let's look forward to the development vision, clarify the current position, and look forward to a bright future.

This article will mainly introduce the following: 1. What is CairoVM? 2. What is Cairo? 3. EVM architecture; 4. Kakarot architecture; 5. Type classification of zkEVM; 6. Kakarot roadmap; 7. Personal thoughts.

Kakarot zkEVM is an EVM deployed using the Cario language, extending the reliability of the Starknet ecosystem by enhancing EVM compatibility. Let's understand step by step.

**1. ****What is CairoVM? **

Kakarot runs on the virtual machine CairoVM, which is the infrastructure of Starknet. The main features of CairoVM are as follows:

· Represent execution as a polynomial equation for provably provable execution.

· Allow all Starknet transactions to use STARK verification.

**2. What is Cario? **

Cario is a Turing-complete STARK-friendly CPU architecture.

· Turing Complete: The system is capable of computing any possible computation/program.

· Stark Friendly: StarkWare's authentication system. Computational integrity off-chain is proven by provers and verified by on-chain verifiers.

So, how does Cario work? Developers can use Cairo to write programs in CaroVM and describe statement statements in a high-level language (to be proven). This improves the development experience, as developers can take advantage of the scalability of ZK Proofs (ZKPs) without having to learn how to write complex circuits.

3. EVM****architecture

The EVM architecture is a mature and well-defined VM (virtual machine) architecture with:

142 opcodes: instructions used by the EVM to execute transactions and smart contracts (SC). A gas fee is required to execute.

9 precompiles: Smart contracts built into Ethereum to handle common operations. Can be invoked with a standard gas fee.

4. Kakarot****Structure

Kakarot is developed based on CairoVM, which is:

· EVM bytecode interpreter

·Smart contracts deployed on Starknet

· Use Cario language

Kakarot supports:

· Deploy the EVM smart contract as it is

Kakarot is not:

·Blockchain

Compiler: does not convert Solidity code to Cairo

As of May 2023:

100% bytecode architecture has been realized (Type 3 zkEVM)

9 EVM precompiled to achieve 8

After implementing all 9 EVM precompilations, Kakarot will become Type 2.5 zkEVM.

5, zkEVM****Type****type

(1) Type 1 zkEVM | Fully equivalent to Ethereum

No changes to the Ethereum system to support easier proof generation.

advantage:

The ultimate solution for Ethereum expansion

shortcoming:

· Computing intensive

Proof time is long (takes hours)

Example:

·Scroll

·Taiko

(2) Type 2 zkEVM | fully equivalent to EVM

Minor modifications to the Ethereum system (different hash functions) make development easier and proof generation faster.

advantage:

· Support most Ethereum dApps

shortcoming:

The inefficiency and unfriendliness of the EVM still exists

Example:

·Scroll

(3) Type 2.5 zkEVM | Equivalent to EVM (except gas fee)

Added gas costs for specific operations in the EVM that are difficult to prove using ZK.

advantage:

Less risky than wider EVM changes

shortcoming:

Reduced compatibility of development tools

· Some dApps are not compatible

(4) Type 3 zkEVM | almost equivalent to EVM

Features that were particularly difficult to implement (such as precompilation) were removed.

advantage:

· Faster proof times

· Easier EVM development

shortcoming:

· Some dApps need to be rewritten

Example:

·Scroll

·Polygon

**(5)Tpye4zkEVM|**equivalent to high-level language

The smart contract source code (high-level language) is compiled into a ZK-SNARK friendly language.

advantage:

· Avoid a lot of overhead

shortcoming:

The contract address may be different from the address in the EVM

May not support handwritten EVM bytecode

· Debugging infrastructure cannot be inherited because they run on EVM bytecode.

Example:

·zkSync

·Nethermind

6**、Kakarot Route Map**

(1) Phase 1 | Bringing EVM to Starknet

Kakarot will initially exist in Starknet as an enshrined EVM. The development experience and user experience (UX) will be the same as Polygon, Scroll or Ethereum.

(2) Stage 2 | L3 zkEVM

Deploying zkEVM application chains through Kakarot allows them to settle transactions on Starknet using proof of validity.

By fusing Kakarot and Madara Starknet sequencer (substrate-based Starknet sequencer) into a unified stack.

(3) Phase 2 | zkEVM-Rollup-as-a-Service

With just one click, rollup receives:

·Application-specific zkEVM deployed on Starknet

· Access to the EVM environment

· Fast execution

Low gas fees: use data availability solutions (Celestia, EigenLayer)

·Encryption security

Advantages of zkEVM-Rollup-as-a-Service:

Run Solidity smart contracts in CairoVM by using Kakarot

Any Solidity smart contract deployed on the EVM can run on Starknet without code changes

Best of both worlds:

Efficiency with EVM

·Smart contracts become provable

(4) Phase 3 | Type 1 zkEVM

To enter this stage, Kakarot must:

·Use Cairo to write Ethereum consensus rules within the full nodes of Madara and Kakarot to prove L1 consensus.

· Switched from Pedersen Merkle Patricia Trie (MPT) to Keccak MPT.

However, Phase 3 is uncertain. It depends on the Ethereum roadmap: Verge. Currently, implementing Keccak MPT in a provable and low-cost way is the main impediment to zkEVM compatibility. After Verge, Keccak may be replaced by Poseidon as the hash function of choice for Ethereum.

7**、Personal thoughts**

Bringing EVM compatibility to Starknet is certainly a huge step forward, but there are still some concerns about Kakarot's success.

(1) Highly competitive ZK field

Competitors are:

· ZK-rollup with different proof systems (SNARK): Scroll, zkSync, Polygon, Taiko, Linea.

·Optimistic rollup：Optimism、Decision、Base

Different zkVM: RISC Zero, Hyper Oracle

(2) Product Market Fit (PMF)

Overall, rollup-as-a-service is an open topic, with two key aspects to consider:

· How many rollups need this service?

Does rollup prefer to build sovereignty and customizability internally?

(3) Continuous product iteration

Kakarot is developing and building a product that is highly technically complex and will likely require constant iterations to be successful.

It also relies on several dynamic components, including:

·Milk

·Data Availability Solution

Ethereum Roadmap: Verge