Rollups for Scaling

Hello, everyone. I am Himanshu. I work for a blockchain startup. In the past, I have worked for some other blockchain startups.

I was reading about the Rollups, and since I am writing a blog each month in 2026 (at least for a couple of start months), this writing will be my understanding/knowledge of them.

Let’s begin!

What’s this thing

Have you done a transaction on the Ethereum blockchain? What was your experience? Would you consider your experience a good one? No one would say it was great. Because you can’t have a good experience on a slow network in terms of transactions per second, especially when you are a popular network, like Ethereum. And, it is not just about the speed of the transactions; it is also about the fees. That increases when the network gets some good activity.

It used to be very bad, but Ethereum has improvements in both transactions per second and the increment in fees, even when the network has high activity.

That’s when Rollups come into the picture, known as Layer 2 networks. They sit on top of the Ethereum, Layer 1. They are the ones that give you what you will call a good experience. They can handle multiple transactions per second. They do not charge a high fee even when the network gets high activity.

How do you do that

What makes Ethereum slow? What is the blockchain trilemma? How are Rollups so fast?

According to the blockchain trilemma, a network can only achieve two out of three: decentralization, security, or scalability.

The Ethereum network is secure because of many validators validating each block, and that’s where they lose their scalability, as it takes time to coordinate among all the validators. And, they are decentralized. Layer 2 networks use Ethereum security to secure themselves, and that’s why they could achieve scalability.

Now, it might come to your mind, what changed? it just a layer above another doing the same thing, using validators to validate each block, and that’s correct, but now, instead of storing each transaction individually in a block, with Rollups, we are storing a batch of transactions.

The next thing that must be coming to your mind is, how does batching transactions improve the transaction per second count? Batching transactions alone does not change anything, but what these Rollups do is, depending on their type, either consider a transaction valid until proven wrong or invalid until proven right.

Rollups can be of two types: Optimistic Rollups and ZK Rollups. The first one considers every transaction valid until proven wrong in a certain period, and the other one considers every transaction invalid until proven right with a mathematical proof. And, this is what improves the transaction per second count, as we don’t need any such coordination.

And, Layer 2 batches transactions so the fees could be divided among many, so they seem high to an individual.

Is that the end

I will end the discussion here. Thanks for reading it to the end.

If you want to go deep, you can find a lot on the internet; this was just a high-level view gist on the Rollups.

You can follow me here to see what I am doing next.

Sometimes, I also record my screen while coding. You can check all the recordings I have done so far on YouTube — https://www.youtube.com/channel/UC1j1TrFcvdHh5HcfFJiBuYQ

If you want to see my projects on GitHub, you can check — https://github.com/hsnice16

You can check my personal logs here — https://personallogs.vercel.app