In my previous post in this series, we went over zkSync and the zkEVM. Let’s go over StarkWare’s ZK rollup solution, which uses STARKs instead of SNARKs.
StarkWare
StarkWare was founded in 2018 by Eli Ben-Sasson, Uri Kolodny, Michael Riabzev, and Alessandro Chiesa. StarkWare has the goal of scaling Ethereum using ZK proofs via StarkNet and StarkEx.
STARKs
Outlined in further detail in my previous SNARKs vs STARKs post, STARK stands for Scalable Transparent ARguments of Knowledge. STARKs have several advantages over SNARKs, namely:
- Trusted setup is not required
- Faster computation times compared to SNARKs
- Hash functions (STARKs) are faster than elliptic curves (SNARKs)
However, STARKs cost more gas to settle on L1 Ethereum.
Architecture
The StarkWare stack consists of several key features:
- Shared Prover (SHARP)
- Generates the proofs and hands off to the Verifier
- saves costs by bundling computations of multiple dApps into a single proof
- Verifier (existing on L1)
- Verifies SHARP proofs as valid. If so, the hash is stored in a fact registry.
Applications either interface with StarkEx (in the case of enterprise and permissioned applications) or the SHARP directly (in the case of StarkNet applications).
StarkNet: StarkWare’s permissionless rollup
StarkNet is StarkWare’s permissionless rollup. Any developer is free to build on top of it and reap the benefits of STARKs. StarkNet Alpha launched in November 2021. As the alpha released relatively recently, there have not been many protocols, dApps, or wallet providers (with the exception of Argent X) deployed, but developer interest has been increasing.
Once StarkNet is fully live and battle-tested, StarkEx protocols can migrate to StarkNet to reap the benefits of composability if desired.
StarkEx: StarkWare’s enterprise STARK solution
StarkEx is StarkWare’s enterprise STARK solution, and has been operational since June 2020. Currently, it serves as the backbone for dYdX Exchange (a perpetuals platform), Sorare (a NFT collectible football management game), Immutable X (a NFT-focused platform, recently partnered with GameStop), and DeversiFi (a decentralized exchange).
Cairo and Warp
With new technology comes new requirements. Cairo is a custom Turing-complete language developed specifically for STARKs. The benefits are outlined in this thread:
While Cairo is better for performance and scalability, it is still incredibly new. Developers have not have had enough time to get familiar with it, and existing L1 applications cannot easily be ported to StarkNet/StarkEx. Warp, developed by Nethermind, is a EVM to Cairo transpiler. Warp allows Solidity contracts to be converted to Cairo seamlessly, allowing developers to deploy on StarkNet.
Conclusion
StarkWare, much like Matter Labs, is building the foundations for an infinitely scalable execution layer via ZK rollups. While Matter Labs is taking a crypto-native approach, StarkWare is building an enterprise SaaS model to capture a large amount of the existing enterprise market. STARKs are newer than SNARKs, but they are already being shown as an effective solution for provable offchain computation.
Thanks for reading this article! This is the summation of a lot of research to help bridge the knowledge gap. Follow me on Twitter to get notified about future posts and let me know what you think. My content will remain 100% free, forever, and is licensed under CC BY-SA unless otherwise specified. Cover art credit: Mr. Hasgaha (Special thanks to Shannon and Swagtimus for the proofread!)