WTF is Surge: The Based Rollup Template - Aayush Giri | Nethermind

Hello. So, so the topic for today's talk is what the is search and what are and how it plays an important role in being a base rollup template. So uh I would like to introduce myself as well. I am Aayush. I work as a senior developer relations engineer at Nethermind.

And so there's this criticism that gets thrown around a lot lately, which is that roll-ups are just fancy multiix and they don't really uh inherit to Ethereum security. And to be honest, roll-ups are still pretty slow. And now I hate to say it uh but there's actually some truth to this. So when you look at most of the current L2s all of them are running on centralized sequencers. That means there's literally one single entity that is responsible for ordering the transaction and also for sensoring sensoring the transactions.

And if that sequencer goes down the entire rollup would stop working. So uh the problem is that we have ended up isolating all these different parts of the ecosystem into separate silos and we do which they uh which don't actually talk to each other in the way Ethereum was intended to do. So we have essentially traded away everything that made Ethereum special uh in terms of uh taking more profits and making more transactions on the chain. So, but the good news is that Ethereum's road map is actually pushing us to fix this uh to build roll-ups that prove these criticisms to be wrong. And this is what Serge is about.

So, if current rollups have all these problems, what should a neutral roll-up setup would look like? What properties should it have to remain true to Ethereum's core values? So first it needs to be permissionless, trustless and censorship resistant which basically means that no central authority would be there in deciding who gets to participate in the chain or which transactions to get processed in the chain. The second point is open source with an open license which means truly open so anyone can fork it improve it or whatever they want to do with it. And then third is anyone should be able to operate this technology uh using themselves.

So if you are able to run an Ethereum node, you should be able to run this as well. And finally, it is it needs to be performant and reliable because what's the use of being decentralized if it does not work in the first place. So these are the properties that built search. So let me introduce you to search. So our answer to all these problems is search again and I would like to emphasize on the word template that we have that we have been using uh for all the slides.

So search is basically a based rollup template that showcases uh nethermind client's gigas performance. Now when I say based I did not say it just to be uh uh just to look cool but although it is actually cool and uh so based literally literally means here which is based on Ethereum. So instead of creating a separate sequencing layer, Serge uses the entirely uh PBS system which Ethereum already has with its own set of validators for transaction processing and which is already part of Ethereum's infrastructure which means surge has these four key properties that solve the problems we just talked about which is stage two at launch based ETH only and gigas. So let me walk you through each one of these properties one by one. First is stage two at launch which means most rollups usually promise that they will decentralize their central uh sequencing layer further in the future and they ask us to trust themselves with this thing but most rollups uh promise this thing to be done eventually but I have not seen anyone doing it till now which means uh censorship resistance security and permissionless functionality is not given to these roll-ups but that is what you get with the surge rollup template from day one.

The second is truly based which means the sequencing of these transactions that are done on surge are done by the set of validators which are part of the PBS system already. Okay. So okay and the fourth is uh sorry the third point is ETH only which means Ethereum is being used for the gas and burn 100% of the base fees. No native tokens, no points, no air drops, just pure alignment with Ethereum's economics. And the fourth is Giggas performance.

We are talking about uh Gigas second per uh capacity through the state-of-the-art Nethermine execution client. So this isn't just another rollup. It's a template that proves you can have performance without sacrificing Ethereum's core values. Now one of the most important things about surge is that we are launching it at stage two security from day one. And if you want to understand what stage two actually means let me walk you through that.

So but what's exactly does this mean? So let me break this down. And so L2 breed has this framework that categorizes roll-ups into different stages based on how decentralized and trustless they actually are. So most of these roll-ups are stuck at stage zero or stage one. which basically means that you still have to trust the team who is running the who is running the chain.

And stage two is the final stage. It's where the rollup becomes fully managed by smart contracts instead of people. And here's what it looks like. A 45day exit window for any protected protocol damages. No emergency powers, which means there are no back doors or instant upgrades that can happen on the chain.

There's limited government governance which means multisix can only cue upgrades and not execute them and users withdraw safety if they disagree with the changes. So the key point here is that stage two the system is permissionless. You don't have to trust anyone or even us. The system protects you automatically and most rollups promise to reach stage two eventually and we are starting surge with stage two. Now let's talk something about that makes Surge really different.

We are completely tokenless. It means that uh it isn't a philosophical choice. It's a practical one. Serge uses ETH exclusively for gas fees and token bonds. No L2 token is required.

Uh you may ask why. So the answer to that is that because tokens create misaligned incentives to the chain. So when the teams have to when the teams have tokens, they have to pump uh these tokens and they start making decisions based on the numbers that go up instead of what actually aligns with Ethereum's mission. So going tokenless eliminates the speculative dynamics that distract us. Uh and there are no regulatory headaches that comes with this.

So it also reduces the overhead massively which means we don't need any token distribution mechanism. uh complex governance systems or teams of lawyers that figure out the regulatory compliance. So we can just focus on building what is important. Now search also delivers what we call gigas performance. That gigas performance per second throughput.

What does it exactly mean? So it means that it is powered by the nethermind execution client which has been optimized specifically for extraordinary throughput. We have actually demonstrated 1 GS per second on the base mainet with version 1.3.0.

How do we achieve this? There are several se several technical optimizations that we have done optimized uh and parallelized inmemory pr pruning parallel storage commit processing uh and there are lots of technical stuff I I would not want to go into that and the result is that the significantly reduced tip of the chain which means faster block times in an overall processing environment. So this isn't a theoretical performance. This is proven uh and the real world throw portut that we have been bringing to the roll-up space. Finally uh let's cover what our multirover security model means.

This is crucial for eliminating single point of failures. We are using a two out of three consensus model which requires agreement from one of the uh two of these three provers systems. Uh we have three diverse proving technologies. SGX based TE SP1 ZK proving system SP1 ZK proving system and why there are three different types you would ask. The answer is that if one gets compromised the other two can still secure the network and the chain can move with two out of three progression.

So this removes a single point of failure in the proving layer. Gives you a higher security threshold than single prover architecture and different security assumptions across uh provers minimize common vulnerabilities. The system has fall tolerance. It means that it remains secure even if one prover is completely compromised. That's the kind of redund redundancy uh you need and a trustless proving system.

So what are the things that you can do with search today is that you can visit the website to learn more about what we are trying to accomplish on search.wtf operate any of the components that comes in the search stack and test it yourself deploy your own devet with a based stage 2 rollup uh you can refer the docs.sarch.wtf WTF to know how to do that. Explore the code and contribute to it as this is what we want and bridge some test net e and transact on search to figure out how these things are working.

Okay. So the next part is search powerups which is uh which is a modular part to the search stack. So power-ups are open source plugins which are designed to apply performance and functionality upgrades to the existing rollup codebase which you have deployed. But think of them as a modular enhancement that can boost the entire ecosystem which you have. The first power is pre-confirmation.

This enables fast user experience while retaining the baseness of the chain. By leveraging Ethereum's validator network, applications can act on transactions before they are fully confirmed, which means a block time of lower than 200 milliseconds in the future, giving users the smooth and the fastest experience without any security compromise or assumptions. The second power up that we have is to focus uh power up focuses on enhanced proving infrastructure. We are working with rollup boost to accelerate proving with multiprover infrastructure. ZK proofs are typically resource intensive.

So the optimizations significantly improve and u the performance while maintaining the robust security model. The next one is block building which means these this boost block production and sequencing speed using the nethermind execution client and rollup boost. So search based architecture creates unique opportunities to builders to optimize the value extraction and coordinate across L1 and L2 creating more efficient markets. Finally, there's seamless composability which means the ecosystem becomes increasingly fragmented uh will become increasingly less fragmented. This provides less frameworks to seamlessly integrate layers of unified liquidity.

applications can work across different roll-ups through standardized uh messaging protocols and the key point is that roll-ups can adopt to these powerups as and when they require depending on the use case that they're catering to. So how did we actually build all of this? What were the key engineering decisions that we made to achieve these properties that we have with search today? First, we use the Tao stack as our foundation. Tao has already solved a lot of the hard problems uh which are around the base rollup infrastructure and we didn't need to reinvent the wheel again.

From there, we made several key modifications. We replaced the finality gadget to work with a multi-rover system, work with a multirover system that I just showed you. And we replace the bond token types to ETH exclusively instead of requiring a separate uh token. That's how we achieve the tokenless design. And I already explained what the features are if we adopt to a tokenless design.

We have replaced the execution client with Nethermind's optimized execution client to get the gigas performance. And finally, we added an extensive documentation so that anyone can understand how to have their own based rollup spin up from the ground. the key points that search isn't a rollup. It's a template. I would like to emphasize again, it's a template that uh these decisions have been made to show you exactly how to build a rollup that prioritizes Ethereum's values.

You can take this forkit and make your own chain uh with your own modifications, maybe even use search powerups to have some more functionalities in your base rollup. Uh so this is about advancing the entire rollup ecosystem and not just launching in another L2. So what does what does it mean when the community takes over the search? It the Ethereum's R1 is basically a new rollup project which is uh trying to uh align with Ethereum's neutral infrastructure values and it takes completely different approach to its funding and uh governance. So it's fully donation funded.

First of all, there will be no VC funding, no private funding, only ETH contributions from the community. No misaligned incentives from investors looking for returns. Second, it's designed as an Ethereum public good, which means the DAO controls all these funds and supplies the engineering team with resources to continue development. All fees fund R&D and application development support until 2030. Anyone should be able to take over this if there's a need for the community to do so.

The model is simple. Whoever donates becomes part of the DAO that participates in the governance of this rollup which is Ethereum R1. And this is about building infrastructure that truly serves Ethereum, not the other way around. So here's the bottom line that I would want to leave you all guys with, which is the future is based. So we are not building another rollup.

We are building a roll-up template for the community to take over and make Ethereum and feel like Ethereum again. This is about being back to what made Ethereum special in the first place and technically not for profits. And um the technology is ready, the template is open source, the community can take it over from there and the future is based and let's build it together. And uh all right, so how do you actually get involved with search? We have made it really simple.

First just visit the website and see the road map and vision that we have. Operate any component. I have already walked you through this and deploy a test net. And uh yeah, so this is the presentation. Thank you guys.