Why 90% of all blockchain transactions will happen on ZK rollup chains within a couple of years

Apologies for the clickbait. Obviously, I have no idea if that's what will happen, but I'm not seeing any other way out. ZK rollups are the be-all, end-all for blockchain tech.

Before we begin, I just want to clarify that I have no interest in tribalism and crypto cults, so if you want to shill your pet chain that has nothing to do with zkRollups, please don't bother. I'd love to discuss zkRollups and their possibilities, of course.

First, a brief on the blockchain trilemma. You can't have all three, pick two: security, decentralization and scalability. But what if one blockchain doesn't have to do it all? What if one chain can focus completely on scalability, while work in cooperation with another which does security and decentralization best?

Enter rollups. A rollup chain is basically a blockchain which is focused on doing one thing, and only one thing, well - extremely high throughput, executing a massive amount of transactions per second. It them compresses and rollups up the "result" of these transactions to the L1 chain. This can result in a 100x improvement in throughput relative to the L1 chain with no compromises. The revolutionary feature here is that the rollup chain need not have its own consensus mechanism - it will rely entirely on L1 for security and decentralization. zkRollups use zero-knowledge proofs to ensure correctness between the rollup chain and L1 chain purely through cryptography. The end result is even though the rollups execute all transactions, you rely entirely on L1 for security. If the rollup fails entirely, for whatever reason, you can simply exit directly from L1. Or just reconstruct the rollup's state from L1 and resume where it left off! Thus, for the first time in blockchain history, you don't have to compromise on security and decentralization.

What if all the data availability offered by L1 isn't enough? What if you want more? ZK rollups can expand things further with further splitting up data availability and transactions. This is called Validium. zkPorter is the best example of this - it has a consensus mechanism, but only for data availability. This way, you can do a gazillion TPS, but still have better security than most chains.

I'll also acknowledge efforts like Mina are trying to achieve the same concept of compression on the L1 level, but the critical flaw here is that they entirely have their own consensus mechanism, which is going to be compromised. I don't see why anyone would use Mina when you can do the very same thing on the L1 with the best consensus mechanism.

Currently, all ZK rollups use Ethereum as their L1, for obvious reasons, but ZK rollups are not tied to Ethereum or any chain. They can and will simply use the L1 with the best consensus mechanism and data availability. My recommendation to L1 chains thus becomes: don't bother competing with ZK rollups - you'll always lose. Rather, pivot to making your L1 the best place to host ZK rollups, or of course, become a ZK rollup yourself.

So, what does this mean for the end user?

You have a blockchain that is fully secured by the whatever the best L1 chain is at the time, but at the same time, you can pay cheap fees with instant transaction confirmations. Indeed, there are further options and flexibility that allow even lower fees should you choose to compromise some level of security.

There are multiple application-specific zkRollups live currently, which you can check out for yourself - zkSync 1.x, Loopring, dYdX, Hermez etc. All of these do thousands of TPS, some of them have such low gas fees that they are effectively subsidized to zero gas by the protocol, and some of these have been live for over a year now. Later this year, we will have the first smart contract rollup chains - zkSync 2.0 and StarkNet. I fully believe this is the greatest innovation in blockchain tech since Ethereum introduced smart contracts in 2015, and expect a vast majority of blockchain activity to happen on zkRollups as these mature over the next couple of years. I don't see any other way out - it simply doesn't make sense when zkRollups offer you lower fees, higher security and decentralization, than any other blockchain solution to date.