Gelato: Should you be running your own L2?

Transcript

Hi everyone. Uh my name is Anerud. Uh yeah, before I get started, uh can I get a quick show of hands of people that know what an L2 is? Perfect. Right. Uh and can I get a quick can I get a quick show of hands if you know what Gelato is or what Gelato does? Cool. Perfect. Nice. So yeah, I'm Anerod. I do developer relations here at Gelato. And yeah, today I'll be talking about should you be building your own L2s. And the reason for this talk and the motivation behind this talk is pretty much a couple months ago there was some FUD on the on Twitter that hey, it cost $40 million to run your own chain. Um, and yeah,

obviously that's that's not the case. But yeah, today I'm going to be walking you over uh and I'll be showcasing you the state of the L2s, the pros and cons of having your own L2s and should you launch your own L2s in general and the future of building web apps. So yeah, let's let's get right into it. Uh first of all, what is Gelato? So we are a rollup as a service platform and what that means is that customers come to us to deploy their own chain and yeah we work with frameworks like the OP stack uh the arbitum orbit stack the avalanche L1 and we have our own in-house stack called the ABC stack which is the L1 sover and rollup stack and pretty much

we don't really just deploy the chain we actually bundle everything together this means that you will get your uh you know your account abstraction SDKs, your embedded wallets, your pay and bundlers, you know your RPCs, everything you need pretty much to to run your chain to have that uh isolated environment that let's say base or ethereum has. So yeah, it's pretty much everything bundled up. Uh and yeah we provide this to you and it's not yeah so basically we have our own in-house uh gelato web services which we call the gelato relay the functions the VRF for uh randomness the oracles for pushing onchain prices

um and then we we do have a set of third-party marketplace as well um and these are basically uh we we we work with 80 plus infrastructure providers as to provide you with the with the tooling that you need and these out of the box solutions that you can have for your chain as well some you know uh let's list a few so like gold sky for example for your indexing needs for example right then you have the block scout for your block explorers then you have hyperl for bridging across uh route scan tenderly so basically we have this set of uh very carefully selected third party infrastructure provider that we work with and you can easily request them to be added uh to your chain as well. Right? So yeah, we are trusted by

over 700 teams. We have deployed over 52 chains. We have over $600,000 in TBL across all gelato chains. And yeah, so currently we are running the chain for Inc. managing deploying it. We also have our infrastructure deployed on Barchain as well. We are you know managing and deploying the chain for rhea which is an arbitum orbit uh L2 chain which is for high throughput for uh oneclick gasless trading and then obviously we are working with Infinex as well. I'll be talking more about that. So yeah built on gelato we're working with Inc. uh they are they built an OP. So basically we we're working

with them on the OP super chain and yeah what what the power is that the D5 L2 ecosystem so that's by Inc. Then we have Fox as well and Fox is pretty much the use case is that they want to prove news articles and pretty much they went with the the CDK. And then we have eBay which is using Gelato Relay under the hood for gasless luxury item authentication B system. So that's eBay and then we have the League of Kingdom which is a L2 gaming platform by Arena Z or not games that's also part of the super chain we are managing upgrading and deploy and deploy the chain for them as well. Uh

and interestingly enough the gaming vertical seems to be quite big. Uh I do have a few stats as well uh in the next upcoming slides but yeah gaming seems to be pretty hot uh when it comes to uh launching your own chains. I'll be talking more about that uh in a couple of minutes, right? Uh a few more use cases. For example, we have Decentraland uh which uses gelato for gasless trading as well. They're using gelato relay under the hood. Uh and like I mentioned, Infinex uh they're using gelato relay for gasless transactions for also with ERC 437 for account creation and switching. Then we have a which is making use of web tree functions for reward distribution and GMX is also

making use of gelato really under the hood for gasless trading and for their perex exchange right so now that we have covered up all the basics about gelato uh let's talk about the state of L2s uh right now right so we are seeing pretty much an uptick of rollups being launched uh you know after like 2023 2024 and then now that we are actually in 2025 we are seeing this obviously going to be more more uh more change launching but if if you see that the gaming vertical has a pretty pretty big chunk of the of the you know the share of the chains

launching so that's a very big vertical as well followed by defi chains are being launched launched and ecosystem chains and all that but gaming seems to be interestingly enough a pretty big vertical and gaming uh obviously it's a pretty big vertical but every major app and ecosystem or an enterprise is launching an L2 so let's take a look at alt L1's for example Lisk for example Astar or LF0ero they were all L1's but now they're recently moved to uh L2s so for example Lisk part of the OP super chain as Alf zero for example now and then apps for example like unis swap worldcoin

frack and rhea they also move to uh L2s then enterprise chains like inc coinbase and fox also move to an L2 and then games for example like league of kingdoms and proof of play yeah they they're also moving to u uh L2s as well and the interesting thing about games is that they don't really so the the reason why they need their own L2 is because they're looking forward for a high throughput and more scalability and it might be the case that they don't really care about security that much from Ethereum point of view but uh yeah they're looking to accommodate all the users uh for their game so yeah that's that's the take so yeah so now that we're here so

like I mentioned initially that the motivation of this talk uh does it really make sense to have your own L2 right so yeah we were seeing over here that you spend either $40 million a year to run your own chain or you spend $0 to run your own chain and build on Sonic, right? Uh so I'm going to be walking you over that and we're going to be arguing against it. Uh and pretty much uh try to break down things over here that it doesn't really cost uh $40 million to run your own chain. And let let's take a look at the cost breakdown over here based on uh these tweets that went out on Twitter a couple months ago.

So here here are the realistic cost given that you want to have a full suite of infra right so you have the chain with basic usages right uh you're going to have the block explorers you're going to have the RPCs right and then you're going to have the Oracle network support then you're going to need like account abstraction services let's say then you're going to need multisc support for let's say account uh smart contract deployment and upgra upgradability then you're going to need the bridging as well apart from the canonical bridge uh the indexings as well and let's say a few automation services. So this full suite of service comes to about close to $20,000 right? So this is $18,500 which is not $14 million right?

It's still expensive but uh it's not $14 million. So uh let's say if if you haven't raised around or so it probably doesn't make sense but if you have actual usage and if you have actual users then it might be worth for you to actually uh you know go ahead and try to launch your own chain. So yeah, there are a few drawbacks as well of launching your own chain which would be that you have to pay these variable costs and what that means is that with the DA and with the settlement when when you're posting transactions uh to back to the L1 or L2 assuming that you're an L3 but that doesn't really makes any you know

there's no significant cost for L3s but assuming you're an L2 uh you got to post transactions back to Ethereum and these are the variable cost uh where you're supposed to pay for transactions. So that's a variable cost. uh then you have this lack of atomic composibility and what that means is that given that you have your own chain and if you work closely with let's say a or compound for example then that it's not it's not really composible because supposed to have the protocol deployed on your chain as well and your users will essentially have to bridge assets to your chain and obviously there will be a lack of native support for on-ramp and yeah there's going to be some you know no native

support for uh the centralized exchanges or the circle USDC for example and yeah there's going to be lack of L1 like grants so there are drawbacks as well obviously but uh for but for the benefits of launching your own chain that would mean that if you have your own chain then you're going to be making revenue from transactions so previously I mentioned that you have to pay uh these variable costs for DA and settlements but that can actually be um you can actually make revenue from transactions, right? And you can actually take a margin for each transactions and from your users and pay for the DA and the settlement costs and

you can actually make revenue from transactions sequencing as well. So that means capturing MBV and yeah, so if you take a look at Ethereum POS, you know, $800 million got leaked since Ethereum. uh and yeah basically if you take a look at revenue from uni chain for example or uh unreal is basically close to the this is the profitability of uh real so it's close to about revenue so the revenue is about about close to 20 $20,000 uh that that's from August 2024 so as you can see that it's going up uh and these are the settlement costs that that that keep

decreasing and basically you can actually customize this as well uh as you move so that's an advantage I'll be talking more about that in this slide which is yeah basically it's your chain as your own rules uh so you can do your own block building as well you can push your oracle prices at the top of each block and like I mentioned you can configure your rollup for maximum security and minimize uh for maximizing security and minimizing costs and this means that tweaking around with the block times, the block gas limits, the settlement layers, etc. Um, obviously there there are a few uh optimized configurations which you have to be aware of but at the end of the day you can configure it yourself. Uh, and

you can decide on bridging as well. The interesting thing about bridging is that um only uh onethird of the users let's say if they have a chain they make use of the canonical bridge which is the enshrine bridge uh given to you by let's say the ops stack or orbit uh but the the twothirds of the transaction of the assets coming into chains for our customers they're actually done by third party uh bridge providers so yeah you can decide on bridging on which bridge should be used it doesn't necessarily have to be the canonical bridge provided u and with this having your own chain you can actually manage or upgrade and decide which feature should be used right so let's say with the current

pectra upgrade for example 7702 that's currently in the hands of the ET foundation to actually implement and upgrade the chain but having your own chain you can actually you know uh implement it yourself I I mean is that that's just an example obviously uh that's uh you know a scenario that I'm just making up. But yeah, that's just an example. And pretty much you can build horiz horizontally sharded apps that live on multiple elastically created chains. So yeah, pretty much your chains, your rules and you can configure it the way that you like. So yeah, who should actually launch their own chain? Uh like I mentioned, you should deploy an L2 if

you expect a lot of users and if you accept a lot of if you expect a lot of me. But given that you haven't raised around and you don't really have any users and you don't really have any money uh then you should probably consider uh going ahead and deploying and building on an existing chain. Um so yeah that's that's the that's the thing and pretty much let's let's go ahead and talk about what if web 3 actually looked quite similar to web 2. So what does that mean? So that means that uh currently we are seeing a trend in the market where launching your own rollups is significantly getting cheaper uh as the time progresses. So that means maybe

over the time you know you can actually run an L2 for $185 a month. Uh and this will go down 100x from over here and pretty much everyone has their own server and think of it like superbase for example where you can just quickly instantly spin up the instance for you know your backend for example. This would also be doable with uh roll-ups as well with your own blockchains. Uh and you pay for usage with the versel like UX. So you just deploy your app uh and in the background you infinitely run the horizontal scalable uh blockchains. Um and yeah another something that that could happen is that the 80% of the uh liquidity would be from the few bank chains and they will

take care of the liquidity provided to to the chain. So most of the chains are going to be app chains and pretty much they use this liquidity uh from these uh bank chains that that are there to provide liquidity to all these apps. Uh and yeah so basically gelato is by uh developers for developers. So to sum it up, you build your own chain. Uh you can actually configure your uh arbitum orbit stack with your own opt L2 or with your own avalanche L1 or the own in-house ABC stack L1. Uh you can set up your own uh DX with all the global RPC nodes. Basically bundling everything up,

connect your chain with the bridging and set up native yield and secure your chain with verifier nodes. And you can extend your app with all these add-ons like uh account abstraction, gelato functions, verifiable randomness, and access to price feed using gelato oracles. Uh and yeah, that's that's pretty much the slide. So yeah, let's build together. Let me know if you have any questions regarding roll-ups. And thank you so much for your time.