Starknet vs Polygon zkEVM?

I think it’s widely accepted by now that 2023 will be the year of zk. But with so many rollups competing with each other it’s hard to understand how will they end up coexisting (i.e. if they do).

I mean, Polygon zkEVM is clearly going to make it really easy for devs to get onboarded. Copy paste the same code from before and it works. Or a few changes at max. Starknet on the other hand has chosen to use a Cairo VM. Now, from what I know, Starknet has managed to build a very healthy developer ecosystem. However, what are the advantages that a zk like Starknet will have (if any) against these zkEVM rollups?

Is Cairo going to make Starknet way more efficient and scalable? Is zkEVM just a hacky fix? Or maybe the other way round, Cairo being efficient is just a myth and the little extra efficiency won’t matter at scale as the tech develops. Sometimes I even think that there will be certain types of applications which will make more sense on Starknet and others which will make sense on zkEVM but I am not sure if this is true either.

Also, I am aware that there is a team building a zkEVM for Starknet as well. However, I feel native support will eventually be better and to excel at Starknet you will need to know Cairo.

View Source

1 thought on “Starknet vs Polygon zkEVM?”

  1. As a contributor on Kakarot (ZK-EVM on starknet) and Solidity developer I think both have an interesting approach. EVM is not ZK friendly by default (not going in details caus don’t know them) so to make a ZKEVM some important sacrifices have to be taken. And on the other side Cairo is built specifically for ZK, in reality cairo is not a language but more like a virtual zk-native processor unit. When you program in cairo you don’t think about ZK. And cairo’s goal isn’t to be a programming language for ZK contracts but for ZK in general so a much larger job offer in the future (we hope). I’d consider zkEVMs as a temporary solution to prove the efficiency of ZK tech.


Leave a Comment