Breaking the Gigagas Barrier & Unlocking the Sovereign Rollup Era | hilmarxo from gelatonetwork

awesome thank you everyone who came to today's presentation um my name is silma I'm from gelato and today I want to talk to you about the ABC stack which is a a new rollup stack that uh we've been developing at gelato and uh it's the first stack that uh broke the 1 gag per second barrier and um it's a new form of rollup deack that we are very excited about which is called Sovereign rollups and uh yeah today I want to talk to you about the motivation behind building it um what is kind of like the reality of rollups today in ethereum um then actually from our experience with working over 50 different rollup customers what do they actually want um and then yeah how ABC was designed to actually address some of these problems that we identified so why am I on this stage talking about this um today well um we in the past like one and a half years we work with a bunch of different teams together um and help them um deploy their rollups um a lot of notable teams that you might have heard about um using all sorts of different Stacks um and uh yeah I think we we used the ZK Stacks the optimistic go up Stacks all of the different sort of Frameworks that you might be aware of um and we look always of course how our customers are using those uh what are the use cases and um do they actually use um certain Services as if they're like like they're designed to be used or not um and uh yeah then of course I will talk a bit about like ABC stack and why for uh a certain user group we believe it's actually a very interesting and compelling um uh use case so just as a reminder for all of us I think there's a lot of confusion often times to like how rollups really work um so a rollup uh at the very essence of it uh is a blockchain that posts its data it posts its transactions to another blockchain in order to Outsource security and consensus around an ordered list of transactions uh that's basically it that's like a common definition of what a rollup is now where a lot of the confusion usually arises is that um in most of today's sort of rollup Frameworks we always have a bridge enshrined to the actual rollup itself so we talk about the rollup as we when we say rollup we usually mean the rollup and the bridge and um why thises this the case is because these rollups were very were designed with like ethereum is kind of like a hub in mind so um when they were first sort of initiated a couple of years ago they was like okay we have ethereum ethereum is the base layer for everything on ethereum we will create all the assets we will issue all the assets and then we will have the sort of network extensions to ethereum where people can Bridge assets into out to again um and we can secure this bridging via proofs fraud proofs ZK proofs you can do um te proofs whatever you like right um and this is really what sort like people think about ethereum l2s today what a rollup is they usually say it's the it's the rollup plus the uh the the bridge and uh what we did is we had the L1 then we had L2 on this L1 and now there are l3s on top of l2s and now there might be l4s on top of l3s right so this is kind of like the sort of H and spoke architecture that arose and um why do we sort of like do this right why do we design it well because as I mentioned uh these rollups were not really thought of as like individual chains they should do their own thing they were like more like Network extensions to ethereum so they they should actually feel like ethereum but not being part of the sort of layer one um back then also we had a bunch of like very complicated Bridge hacks and bridges where like custom build for every single L1 and there were a lot of issues with that so enshrining this bridge into the system kind of like made sense um and back then literally 99% of all the assets were issued on ethereum also the stable coins right um and so yeah thinking that ethereum is the up for all the assets also made a ton of sense but yeah let's look at some data here right this data is directly pulled from like L2 beat um and yeah what do these sort of like layer twos that you see right here have in common ZK sync Bas real which is real world EST chain arbitrum and so forth well all of them actually have um the majority of the assets on their chain natively issued on the rollup itself so these assets the the the biggest chunk of the tvl of these chains actually do not come from ethereum anymore even though like this was like how most of these things were designed at the very beginning these chains have the highest percentage of the tvl um issued on the Chain themselves so like on base it's mostly like usdc for example and the issuer is actually circle on base and not circle on ethereum and then bridging into base it's like directly issuing on a base uh and there are some where this percentage is over 90% now this is another set of chains like Rayas for example perex that is working together with on so Bob is like a Bitcoin L2 also on ethereum mode Inc is also sort of L2 is working together with us so what do do those have in common well the majority of the assets for those chains is actually bridged in from other chains not ethereum so they don't use the canonical bridge that is enshrined uh to ethereum to bridge over these assets uh they Bridge them over from all sorts of different other chains maybe Solana maybe Bas directly maybe um arbitrum directly so these well tldr they are using these rollups quite differently than they were designed if you look at this and this is kind of like a snapshot of like the entire sort of rollup space you see that for all rollups that exist or for all l2s that exist 47% of all the issue assets are already natively issued on the chains and not on ethereum anymore e theorum is not the asset issuance platform that we thought it would be at the very beginning like did Dynamics have changed and what I basically sort of want to uh sort of make you think today is okay if that's the case if already 47% of all the assets are issued on these rollups could you make the case that these rollups today are already 47% Sovereign and actually not ethereum l2s or they are only to a very small percentage 38% still ethereum L2 this is kind of like how I sort like want you to think um in this presentation so we analyzed um a lot of chains we work together with a lot of these chains right and when I talk to them when I asked them okay what are you actually looking for in in building your ecosystem your application your blockchain um they usually come to me with like five things one of course very simple I want extremely high throughput I want low cost right and even today um a lot of the the chains that exist uh do still have pretty limited throughput uh all also transaction fees SES go up especially if ethereum data availability uh ethereum blob space is filled up then the second one is pretty important it's they want to onboard users very easily to their own ecosystem um and they want to do it fast and cheap um and the problem is every single time you need to force someone to go VI ethereum to join your chain um you have to pay pretty hefty fees and it becomes quite slow that's why a lot of new chains actually want to onboard users via existing rollups or existing other layer uh um chains set base or ink for example then of course you would like to have access to like all the liquidity in the world you want high security low cost um and this is also interesting you want more control over your chain this control means you want to control how transactions are getting sequenced you want to control how blocks are being built um and there are actually applications and um organizations that want to have control over who can transact on their blockchain right uh ethereum a lot is a lot about censorship resistance so everyone can always transact you can never stop them right but if you think about like companies actually like Building Products that want to interface with uh web free then they need some sort of permissioning because they have to be compliant with a lot of things right and they they might not want to allow the Lazarus group to execute some transactions on their on their rollup especially if they are the the biggest application on top so are these sort of L2 assumptions that we set out at the very beginning still valid um well after like talking to a lot of developers I think and it's sometimes the the brutal truth but I don't think they care that much about scaling ethereum they care about scaling the application right ethereum is like a platform they can use in order to achieve this but it's not about scaling ethereum per se anymore it's about hey I want to create a successful application in the space um also like bridging changed a lot like back then there were like ronand Bridge hacks and so forth but these days you have pretty solid like interoperability cross train mastering protocols that have a variety of different security modules that can be ZK optimistic optimistic ZK uh or just like attestations like hyperlane layer zero that are pretty robust if you look at like the tvl and and sort of through P support um then uh users actually on board mostly via stable coin especially like usdc and usdc is still issued on ethereum but it's also natively issued on base arbitrum Solana and so forth so if you want to onboard users you probably don't want to go via ethereum you just issue uscc straight on base or on Solana then you bridge from Bas and Solana to your chain and buam you got your user onboarded why would you go by ethereum it's slow and expensive and the and it's not more secure because uh the settlement layer for UC is Circle service it's not ethereum it's not a native ethereum asset um yeah and so this is this is all like what we uh came up with after looking at all of this um so we believe that there's this world where there's there's this big industry which is like we want to scale ethereum and we want to build this sort of like Network extension layer on ethereum which um uh there's a lot of traction already and I think it's it's it's working very well U but I think there from talking to a lot of customers I think uh a bunch of chains a bunch of projects actually do not want to scale ethereum they would rather just build and have control over their own blockchain that can interface with ethereum but ethereum is not different to like a Solana to them or like even a ton or like a Bas it's it's so like all equal and this is really what the ABC stack tries to achieve it it tries to be this like it tries to cater to this audience um and it does so at a at a really impressive performance that I will go into in a moment and so the design pH opy uh of ABC what it is is it's an abcentric stack it's it's it's not an ethereum Centric stack so rather having this rather having this application sorry ethereum in the center of like your solar system you uh have your application at the center and and ethereum is just a very large sort of Planet hovering your own Planet but it's not different to any other planets like Solana or or Bas or ton for example uh you interface with them all the same um it's much more like a mesh Network design that that it is like a h and spoke uh you can think of more like the cosmos sort of like architecture and design pattern than like the sort of like L2 L L1 L2 L3 pattern um and what it focuses on really is it basically completely gets rid of an enshrine Bridge you cut it out completely you don't have that anymore there's no opinionated way to settle to specific chain and Define your rollup based on some light client that you enshrine but it's you it basically it's just hey you bring the interoperability system you want you can bring a hyper Lane with like an attestation service you can bring layer zero that secures your chain with Google you can bring Circle uh if you pay them a lot of money right it's completely up to you how you want to interoperate with whom you interoperate um and uh what's the security you would like to have now um we have some pretty good performance on the stack so uh it does 1 GHz per second with a throughput which is sort of like 100 times uh um that of like most other sort of like uh technologies that are currently available uh it is very much based on Celestia uh why Celestia Celestia uh as a DA layer has uh the highest throughput already today has one single slot finality which is great for inoperability and basically from a road perspective will already provide capacity that is way way higher than what ethereum will probably achieve in the short to medium future um though of course I still want ethereum to scale a lot and I'm rooting for it but yeah I think they have to increase the pace a bit to to to compete against guys like Celestia here uh there's this nice chart with the banana Zone on the top right this is like the Giga gas banana Zone which ABC stack is on um and yeah you can think you can you see like versus Space today what type of frut is possible um so you can really like bring everything on chain that you want this goes into the modular modular interoperability as I said ABC has multi-chain settlement so you you don't settle to an like you don't have an enshrine settlement to ethereum which finds Z rollup you just have a modular androp ability protocol you can settle to six different chains at the same time uh but like you are still like a chere Sovereign chain and how you want to deal with bridging is completely up to you um and yeah so this is I think how we we think of like rollups in the future they they want to directly interoperate with all these ecosystems um and they don't just want to be like connected to one of them and then and then have this be the asset issuing platform you want to have assets that are issued on salana all the meme Co should come to your platform ton who knows if they if it will still be successful in a couple of years but if so you probably want to be in toate with it right so this is kind of like the architecture more that that um ABC Embraces some other really cool features that um uh it has that I think is very powerful based on sort of customer feedback that we heard uh one is uh custom block building so you can Define your own block building rules uh on the stack uh you can avoid me you can update Oracle prices at the top of the block or pretty interesting for Enterprises you can uh say certain only certain users can actually transact on the chain and can Bridge Assets in and can Bridge assets out we don't want anyone who doesn't comply with these rules to interact with us and this is actually how I see the future we have we'll have all sorts of companies deploy their chains they all have to adhere to compliance on like different countries and so forth but you always have these neutral uh chains in neutral layers like an ethereum or or other chains to still go to and then uh enjoy full sensorship resistance so I think they are just like different sort of products different different use cases uh then you have a native gas token capability and this is not a custom gas token so this is not like a token that you Deploy on ethereum and Bridge in and you can use it for gas this is a token that you issue on the chain and use for gas uh natively so it's like e it's like ether and ethereum it's it's it's truly native asset um then of course it's one to one evm compatible so so no changes at all and it's um ZK forall compatible uh so you can just like click a button you can upgrade your bridging security to ZK um right away uh if ZK becomes fast enough and cheap enough which it is not today unfortunately uh we ran these things in production uh but I'm pretty pretty hopeful that uh with the work of all these amazing teams like R zero exct and and all the other guys that we will get there sooner rather than later so um uh yeah it's it's already forward compatible with it and just like as some examples here how you can sort of take the stack and make it yourself if you have a defi chain let's say how would it look like you have da Celestia you have interoperability use like the hyay messaging protocol you want to have high security you have 500 million in the bridge well use exin or resero to verify uh bridging transactions by ZK and you probably want to be connected to Solana arbitrum and theum because they the highest tvl chains in the evm ecosystem right let's say you're a gaming chain should have gaming chain be like ZK right now probably not right probably it has only 500k in TVO probably it just wants to Min some weirdest nfts so I think having like a three or four multi6 sort of like attestation service to bridge these Assets in an out out of this chain is totally fine right if you can cut costs down and if you can get latency down of like bridging in and out which is what uh games want perfectly fine right and you probably don't want to be connected to ethereum you probably want to be connected to Ronin or B or some other ecosystem right yeah and like finally like Enterprises as I already mentioned you probably be want to be off compliant you want to probably Blacklist Lazarus you want to not have them be able to interact on your chain right uh this is all possible and you probably don't want to post your data to any public da layer you want to keep it private you might want to use CK to verify that everything has been run correctly but that's it really yeah so ABC stack is now available um on our Ro Business Service platform among uh the other like uh leading uh rollup Stacks in the industry uh you can actually literally click it in 5 minutes so like you come to a website you make click a button 5 minutes you got the chain like the fastest deployment in the industry uh I guarantee that yeah so uh ABC is really for building everything in on chain and how you want to and not being opinionated uh based on sort of some Legacy um uh way of thinking about you as an application or a project if you want to build uh come to us come to me after the talk we have a booth as well um yeah we would of course love to support you thanks a lot for for listening in deis