Ethereum is the most resilient, censorship-resistant DeFi infrastructure that exists. For it to support the financial transactions of billions of people someday, it must be more effectively scaled.

The Merge helped with this. EIP-1559 helped with this. And Optimistic Rollups (ORs) have been carrying much of the weight lately, offering an imperfect-but-functional solution to batch-transaction scaling.

ZK-rollups (ZKRs) are the holy grail of this evolution. With about 4x the throughput of ORs (2k TPS vs 500 for OR), no reliance on honest validators, no withdrawal holds, and cryptographically impossible to post a fraudulent transaction, ZKRs are the conduit to take DeFi to widespread, institutional-grade usability.

We had our choice to build on whatever chain we wanted. We made a very intentional decision to do so on Starknet’s ZK-STARKs. We’d like to explain why this represents the best, most-durable choice among ZK-focused chains.

The core tenent of ZK tech is a cryptographic method by which one side proves to another that a given statement is true, while not disclosing anything else while doing so. Said more technically: ZK tech is a proof architecture where one demonstrates specific knowledge, like a key or account balance, without interacting with either the prover or the verifier and sharing nothing else about the knowledge itself.

Let’s demonstrate this by way of entertaining example:

Where’s Waldo? You don’t know where he is. But I do. How can I prove this to you, without showing you where Waldo actually is?

Imagine I slid a white piece of paper over this picture with a small hole cut in it, and I shifted the hole over Waldo to show I know where he is, but obfuscated where he actually is on the map.

Now you see I do in fact know where Waldo is, but I haven’t given up any secret information about how you can find him yourself. I prove I know something without giving up details on anything else. This in its essence is what ZK tech allows us to do.

However not all ZKRs are created equally. There are currently two types of ZKR tech: ZK-SNARKs and ZK-STARKs. The core differences between the two rest in their setup, attack resistance, and scalability.

SNARK stands for "Zero-Knowledge Succinct Non-Interactive Argument of Knowledge”. They operate off a private/public key pair (ECDSA signature) and must undergo a trusted setup phase where we assume developers don’t manipulate code or create vulnerabilities. This trusted setup is a core characteristic of SNARKS. It’s only a one-time event, but nonetheless, you do have a non-trivial exploit opportunity.

STARK stands for “Zero-Knowledge Scalable Transparent Argument of Knowledge”. It performs the same core ZK function of SNARKs, but facilitates it in a much more secure, robust way.

STARKs improve on ZK tech by avoiding the dependence on a trusted setup. STARKs use publicly verifiable randomness in lieu of a key pair for its arrangement, removing any reliance on trust. In addition to being more secure, it also makes STARKs quantum-hack resilient.

There is a small but noteworthy risk that cryptography may come under life-threatening attack from quantum computing. A quantum computer could theoretically break the private/public key pairing by brute-force computing guesses until it found a match. Since there is no key pair for STARK setup, the protocol is quantum proof. When we say we're building with security and the long term in mind, we mean it; the quantum resilience and trustless setup of STARKs is aligned with this.

Additionally, STARKs are much more computationally scalable by speed and size than SNARKs, by roughly a factor of 10.

So STARKs are more robust to attack, have about 10x more throughput, and have no reliance on trust to coordinate. This is why we consider STARKs the pinnacle of scaling technology, enabling developers and users to confidently relocate storage and compute offchain.

We wanted to build RabbitX on the most scalable, future-proof tech possible, and ZK-STARKs represent this. We feel it’s only a matter of time for the market to recognize this. We chose Starknet because they’re the most professional, competent, and technology-minded team we’ve had the pleasure of interacting with.

Best tech + best team = best place to build

Follow at @BackTheBunny

Check out another one of our popular posts --> Predicting Recessions: What Yield Curves Tell Us About the Future Demand for Money