The Future of Blockchain Scalability: Layer 1 vs Layer 2 - TOKEN2049 Singapore 2024

all right good morning everyone uh shout out to all the Troopers who made it out early in the morning today uh we're talking about one of the most important topics I think in crypto today which is scalability we're looking also at specifically L1 versus L2 scalability but I think we'll go wherever uh we'll go wherever the conversation takes us so we got three awesome people on stage today we've got Nicola from eternity we've got Merck from Cello C and then we've got anog from ail why don't we just uh start from the top with some intros um what you're working on maybe some thoughts on skillability to kick us off and we'll go from there so my name is Nico stov I the chairman of the eternity foundation and I picked that up I think about nine months ago in order to focus on Reviving an L1 and preparing it for what is to come and probably also for what we're going to talk about today hey guys I'm Mar kfki I'm one of the co-founders of cell um cell is an evbm compatible L1 today that's transitioning to become an L2 tomorrow so I think this topic is definitely near and dear to our heart uh and if you're new to cell where um again an evbm chain really focused on real world use cases there's a lot of stable coin activity happening on Cell a lot of payment activity a lot of savings activity hedging against inflation um but also more recently uh micr lending and just lending against future income um all of the things that you'd expect from kind of a chain focused on really providing utility for people um you know like the kind of that mirrors is the banking industry but fully onchain hi my name is anur I'm co-founder of aale previously also founded polygon um aale you know like actually was started within polygon back in 2020 we spun it out last year um ail really you know like is infrastructure for uh what we call the model of future of blockchains essentially um the initial thinking for blockchains was that we can scale on one blockchain like lay one blockchains but uh what we think is you know like um the entirety of transactions on blockchains can't exist on one blockchain and so you really need this multi-chain future and so we build a wh is this infrastructure which consists of this layer one that supports rollups on top really and you know like does scalability interoperability and all that stuff uh so yeah uh happy to get into more details later on awesome and I'm Joe from Alchemy uh web 3 developer platform um we got a lot of great perspectives up here I'm really excited to see kind of the different takes on on scalability um cool well why don't we start with the basics I'm curious to hear from each of you guys what do you think are the main challenges around scalability today and what are you personally excited about working on the multiple things that I'm excited about working on but right now I think the most interesting thing is and we had this discussion a little bit earlier with marrick was that we have been working at The Eternity Foundation about on something which is called hyper chains and on hyper chains you would be able to create scalable insecure side chains which you would then be able to underline with for example Bitcoin to make sure that they're absolutely tamperproof and you would be able to utilize the infrastructure of Bitcoin and then on top of that you would be able to play with on a so-called child chain with anything you would really like to so you would combine the proof of work consensus with the proof of State consensus which might actually be one of the things that I'm most excited about and the other thing that I must actually admit was what uh I spoke with anoro co-founder a couple of days ago when it comes to what a veil is doing right now with modalizing different components of the blockchain this also sounds really cool to be honest cool yeah yeah I think on our end obviously the the upcoming transition to the L2 is obviously all consuming and something that we're really really excited about working really hard on um it's no easy feat to turn a you know a decentralized running uh L1 that has on the order of 700,000 daily active users um you know billions of dollars of value being transferred every month uh to to safely and securely without incident trans transition that into an LT too and so that's definitely taking a lot of our time um but um it's going well um we already launched a test net uh a few months ago at ECC and we're looking to transition another one of our test Nets end of this month and then hopefully main net by the end of the year um beyond that you know I think I'm very excited about I think a lot of the work that's happening in the broader ethereum ecosystem um around interop I think interop is going to be the next um really big um I would say bucket of work that the whole ecosystem is working on um you know I think you probably heard a lot about um also solving the ux issue for users who are on multiple l2s uh I think all of this feeds into interop and so that's definitely a big bucket that I'm super excited about yeah so I mean I think maybe I'll try to put it in a more um web to sort of analogy uh like so we looking at scaling these public blockchains um if you look at like how apps on the internet have kind of evolved over time right like I mean you don't have all the world's apps um on the internet live on one single super computer right they are on separate servers they're scaled independently based on demand um and the L1 the layer one blockchain uh situation is something uh of this super computer sort of analogy you know I like to um say it like that right like if you look at 2014 when ethereum was came in and all the iterations after that I like to look at them at attempts attempts at better versions of ethereum so if it's like a phantom BC Solana all of these are attempts at better versions of ethereum but the reality is that all the world's apps cannot fit on one single big supercomputer and so you have to get into that server ification mode mode and based on you know like what we've studied the industry has iterated on the rollup construct on um you know like the um ethereum or um you know like aale for example is this the closest analogy to what we have as servers in the web to in the traditional Tech right like rollups are really these verifiable servers like can we do some compute outside and then provide some input um back to uh the Bas layer that U you know helps in verification for example and that's how you kind of scale and the beauty of this architecture is that rollups or you know rollups are don't need to be evm svm they can be anything like we seeing uh rollup Stacks like stacker for example that are you know you can write programs in JavaScript in typescript in Python for example so I think what I'm trying to say is that on a meta level like this layer to debate is really the server ification of blockchains right like I mean uh how can we get to an architecture that is that allows us to scale uh based on demand like rather than uh you know like scale like um scale your superc computer or your server cluster for example right like so so I am excited about this and this has been already been there for now 3 to 4 years like it really began in 2019 2018 when rollups were starting to getting discussed and now things have started to kind of reach the next level we already have like hundreds of chains um uh as rollups and so the the things what we are excited about is I think the technology is sort of they in place I think what we are more interested in is how can we kind of instead of like we've been this this kind of situation where we we build this very nice systems and we expect users to come over on RAM jump over hoops and like I think the next stage and which you can already see with something like ta is that you're taking the blockchain technology to users rather than expecting them to come here and I think that is going to accelerate with this whole Layer Two sort of a thing where we will be able to embed blockchain technology into user workflows into distribution flows rather than the you know like the inverse I think that's really well said and I think one theme that I heard from all of you guys is really like rollups l2s um obviously it's a big part of the ethereum road map and how we think about scaling today um am I'm curious to dig in a little bit with you so C you guys started as an L1 um but then decided to transition to an L2 how did you think about the trade-offs of that yeah great question I mean for us I think it was more of an ideological decision we started in ethereum we started building an easyto use wallet on top of E um and when we realized we couldn't get it to scale we couldn't get it to have the experience that we needed to to Target those normies um that we just talked about um we ended up working on our own alt L1 evm compatible L1 and that was what you did back then this was 2018 2019 um the scaling road map didn't exist now that it does exist um I think it's very hard to be in the etherum ecosystem and um have an evm chain that is not an L2 right I think for anybody who is working on or running an evm compatible L1 I would sincerely recommend just taking a a deep look at the project and thinking about whether or not it makes sense to uh migrate to an L2 I think ultimately developers um building on on these evm chains will want it um I think from an interrup perspective given all the really exciting things happening in ethereum on interop um it's just going to be a big win I think ultimately what what I'm seeing is um a lot of these other evm uh compatible l1s are really taking a close look at kind of following solo and and doing the same and um and I think ultimately um that's good news for ethereum um I think you know the future is certainly bright for ethereum becoming the settlement layer of web 3 yeah so very focused on ethereum there um Nicola curious to hear your perspective because ethereum and evm compatibility is not as much of a focus for you guys listen I I like building l1s uh this is something this is how I've started off with and this is something that I feel incredibly comfortable with and I believe that it is a vital part of the web 3 ecosystem uh the beauty of layer ones is that you do have heavy infrastructure so it is it is really a Basel but it also comes with its own issues obviously so there's a trade-off on one side I am able to provide technology as a baseline which people can build up on in order to allow different scaling opportunities if like you said anarak if those are ZK or optimistic rollups that are now popping in or if this is with you as an L2 but you still need an underlying technology to provide that stability and security like we just mind our I think millionth block that has we have never been down there has never been any issues and that is in my opinion something that gives you trck record in history so with having that base layer I'm absolutely for this call it maybe symbiosis between layer ones and layer twos because having that base layer and building on top of it with an L2 that now gives you the opportunity to take care and really solve the issues that you have with um with l1s cuz honestly speaking an L1 is not built to be infinitely scalable because you always ethereum was a perfect example for that you had the congestion you had the issues you had the hex Etc PP but with the l2s that are popping up now those things are resolving those things but they can still go back to that layer one kind of deep stable part and make sure that they can rely on it and that secures also their chain at the end of the day on the other hand if your L1 somehow has an issue that automatically then translates also to the L2 so there's again a trade-off into it but this is why for me it's vital that those things coexist and this for me it's incredibly important to make sure that you have different l1s to be able to cater to different use cases because I think you mentioned that earlier on Ro that there's not one chain to rule them all you know seven rings and the load of Rings should be more chains than one to do everything else but do they need to be l1s could they all be l2s well why you can't run everything on one L1 you course you can why you still like with ethereum we saw that they're the same issues that ethereum created or had or came out of ethereum they kind of translated also into the l2s you still have solidity which is programming language I'm sorry you still have uh hacks most used language true javascri most language you know so it's still it's that tradeoff that I'm talking about so ethereum was that front runner they like everybody used ethereum because it's the first smart contract language out there right it was the first blockchain alternative to bitcoin that became let's say somewhat prominent right but I still think that there should be other chains to support each other and to work that's why I believe that interoperability between once is going to be an incredibly vital part of how this continues do you mind if I respond yeah yeah no go ahead this is fun very much this is fun we can't all have the same opinion right I think this is really this is really lovely and and I think actually my response might dovetail nicely to what you're thinking uh on the topic of um can you have you know infinite L2 scaling up to uh 1 L1 you know the answer is Absolut absolutely right because at the end of the day um the main bottleneck is data availability and and as um you know Avail and others have shown you can scale data availability horizontally um and so um the end state will certainly be you know a single L1 um that can scale data availability whether or not every uh chain will roll up to that you know I think remains to be seen but certainly it's not a technological issue um so so and then the question of course is which chain will be the one that everyone's rolling up and and obviously ethereum has um is in the pole position and certainly one that that we're really excited about at so I think uh when we you know discuss this layer one layer two thing uh sometimes we get a little bit tribalistic in general but the idea is that like if you kind of divide the functions of a layer one and Layer Two especially you know like in a very model construct is that like there are two parts to building a blockchain one is the actual execution part which is like Innovation on the VM whether it's the evm svm move VM JavaScript VMS you know like you take it the other is the more uh difficult to pull off thing which is the cryptoeconomic security bootstrapping part like the every time you kind of come up with a new layer on blockchain you have to go through that cycle you have to bootstrap a set of validators you have to bootstrap a token to you have to ensure that a token has value I mean so it's a lot of work doing that over and over again for new layer one blockchains and that happens every single time uh and so really when we kind of talk about this layer two it's we need to also talk about how can we you know like the layer two thing explicitly doesn't need to do this cryptoeconomic bootstrapping uh thing like if you look at Bay from coinbase like what Bas did is they took the op stack code and deployed as a rollup on ethereum and they didn't have to do all this bootstrapping of validators or or even a new token for example they focused on you know like what they wanted to which is user acquisition user adoption that sort of thing and you will see that happening more and more right like so this cryptoeconomic security bootstrapping which is the more slower part right like the token takes time like you you need time to build up this cryptoeconomic security that's why you have all these uh cryptoeconomic security as a service sort of projects like igen layer like symbiotic popping up because that is the really hard problem of bootstrapping a new layer one blockchain um it takes years to like ethereum has been there for 10 years um you know like that's that's how it kind of got to this stage in terms of security and it's very hard like if you can you can look at the cosmos set of uh L1 chains to see uh what kind of cryptoeconomic security they've sustained over the years uh it's very hard um and so that that is the foundational um you know like um lens which you should also view layer TOS The Innovation that you can bring to this space is that it's not like of course the evm is dominant but you don't need to stick to the evm like you can you can do whatever Innovations you want like we have now Solana VM rollups coming we have move VM rollups coming you know like so all that sort of innovation still stays even in a layer to kind of a thing but you kind of solve this from a more scalable of course it provide you know brings in this fragmentation issue of how do you connect all of these togethers but that's also there in a layer one if you have 100 layer ones they still need to interoperate right like there still an interoperate problem but but if you are on this shared Bas layer it's much easier to interoperate right like so so it's much more nuanced than we like to think yeah yeah absolutely and then so you know rollups obviously kind of uh seems like where the industry is going what are one question in mind though is like what are some of the challenges that you guys see with RS I think a lot of advantages like you mentioned bootstrapping the security on the L1 the shared shared kind of uh the shared ecosystem shared Developers but what are some of the challenges that you guys think are there I would go Mar go ahead for me the the biggest challenge certainly is around um you know wallet abstraction and uh and just interrupt right like I think we've basically proven horizontal scalability now of chains which is great um so that means that we we can effectively scale ethereum infinitely um that's that's quite lovely um but now everyone's saying okay but you know suddenly my experience is a lot worse I have um accounts in multiple chains I need to have wallets that can make sense of that I think long term what we really want is we want um what people are calling full chain composability It's ability for a smart contract on one chain on one L2 to be able to interact literally do a function called synchronously not async into another function on another L2 that is a Holy Grail I mean once you do that you basically solve the scalability trilemma and what's exciting is that it's actually feasible I would say for the first time now um there is a a proposed solution for this problem that that you know everyone really thinks is going to work and that solution relies on ZK EVMS and what people are calling um realtime proving so this ability to prove your transaction um in in within the time span of an ethereum block so say less than 12 seconds ideally somewhere around 4 seconds if you can do this then you can actually prove the validity of your transaction with respect to the latest State on ethereum and once you can prove that then you can automically execute another transaction on another chain as part of that first transaction and that effectively allows you then to create function calls that are synchronous between two l and so you know we have ZK avms today um the some of the best uh ZK avms are able to prove transactions on the order of 15 seconds with that kind of latency um and so we're actually not that far away from being able to get to that 4C time period where we'll be able to actually prove these things in real time and you know there's a lot of work of course to to get these zms to be um um battle tested and and secure uh and then even more work to create standards around how uh all of this interrupt is happening in the broader ethereum ecosystem especially because there's multiple zkm implementations out there but I think ultimately you know we're in a position now where um we can actually solve you know really the the scalability trilemma which is just really really freaking exciting yeah I mean um uh see as the modular road map has for ethereum has improved right like I mean in general we also sometimes forget that ethereum was born at a time when the only way to build blockchains was very monolithic it began Life as a monolithic blockchain then slowly over the years it has been you know kind of on this upgrade path and now because ethereum is very uh it holds a lot of value uh on production it's it's of course necessarily slow to upgrade in general and so um some of these things will come soon soon some of these com later for example but the path uh to the end goal is becoming more and more clear and that's why you know like blockchains like aale for example you know we started um uh to fill fulfill that need in general um some of the challenges that we see are you know like things like fragmentation for example right like there's hundreds of chains no way for these to interoperate and it's not necessarily that there's these 100 chains they need to interoperate it's also actually looking at developer problems like something like an a for example is deployed on probably 12 or 14 Chains It's the same exact protocol deployed on these multiple chains they have to maintain these deployments on these multiple chains because they don't want to lose out on users on all these chains how can we get to a position where you deploy a protocol on one chain and it is accessible to users across These Chains like those are the questions those are the challenges that we we have to to answer right like I mean in general like sometimes we get into a lot of lowlevel features but you know how do they translate to the end developer end user in general uh how do like all the wallets that we see today are very power user ux oriented right like I mean you have to the each of them you have to switch chains of course this is changing pretty soon but we need like more uh better wallets built from the ground up for example um like from developer point of view right like I mean if so we started building like this polygon POS chain back in the day then we moved to ZK AVM but if you kind of think of it the developer interface is still the same it's still the evm like there's no real a difference in terms of you know we can bring more compute power bigger blocks Etc but how can we make it easy for the average traditional web to developer to come to this space how can we enable writing these programs in let's say typescript or python or JavaScript but still be able to access these same security guarantees so I mean I think we do a lot of thinking on you know like infrastructure and how one infrastructure will interact with this but we still we as an industry we have to kind of think more and more about how the developer experience can improve how the user experience can improve and more importantly how can we kind of take this technology and put it into the user workflows to distribution workflows uh I mean I would argue that is the only reason why the TA ecosystem something like the tonic system is being you know like becoming very popular is that there is this clutch of users like I don't remember like I don't know if you guys remember there was a time in 2019 2020 when Facebook was coming up with Libra and a lot of the industry you know like people were you know kind of threatened so to speak in general right like I mean the reason is Facebook has 2 billion users I mean if you bring uh programmable ledgers to that kind of audience you know like it's game over for a lot of projects for example right like so so we have to be very clear that this in this technology people will not come to this industry because it's decentralized it's you know like uh you know like you people don't uh buy into privacy tools because you know like it's a very Niche audience you have to go to what the user wants you have to kind of build those distribution flows so those are some of the challenges that I see got to build build for the real world yeah absolutely I'll give it a perfect example when we started building eternity we opted to use um erlang as a programming language so erlang is object oriented program language which is incredibly complicated there's like three and a have people in the world averaging the age of 60 that actually were working for Sony Ericson that use this language the beauty about earling was that um it's literally made for transacting information in a very very secure stable and very fast and efficient way but then we got to the problem that I think you both guys addressed quite well that how do you actually get this to the masses when there's it's so complicated to use this and then we get to in my opinion also I think both of you guys mentioned this the two biggest issues that we have is number one is interoperability how can we all work together in order to cater to more people and the second thing is the user experience in itself because if it is more complicated than something I'm used to it's such a high barrier of Entry that I'll off for something else it's just that simple and lowering those baras is probably the biggest challenge that I see yeah makes a lot of sense cu everything else was mentioned already but that W I think well well said on all those things and we've got time for just I have one more question for you guys so looking 5 years out what does the world look like in terms of scalability does it kind of look like today with rups we've made some ZK improvements kind of solid things other radically new architectures like what does that look like and we'll have to take this one fast I'll go first in that one I honestly think that uh Larry twos are going to become like the main user Focus the main the main point of access for users l1s are still going to be there going to be effective they're going to make sure that again the Baseline is secured the infrastructure is stable but besides that because building layer one and just advancing layer one takes so much more time uh rather than just using all the advantages of a layer two you're going to have like this those bricks and blocks are just going to move on so you're going to have an L1 is going to create something new and then l2s are going to jump on it and they're going to advance way faster so it's going to be a Brick by Brick building part which in my opinion is still going to hopefully get us closer to that nice and easy ux that I mentioned earlier yeah I touched upon this before I mean for me the future looks like thousands and thousands of of l2s rolling up to etherium with full chain composability so all of these l2s will be able to uh call Smart contracts across chains basically have one smart contract be able to call a function in another smart contract running on another chain uh and we'll have all of that uh working with uh real time ZK proving um and some really elegant um you know transaction bundling Happening by by block Builders so yeah and I think in effect we'll have solved the the scaling trilemma at that point see I think at the Hard blockchains are these uh secure programmable ledgers essentially uh and like the world runs on ledgers every workflow in that um in the world is either like a social uh Ledger transaction or an economic Ledger transaction and so essentially in five years I would say that you know like blockchains make their way into real world workflows like they kind of supplement or replace uh The Ledger interactions that happened right like everything in the real world is a ledger transaction your banking is a ledger your social media thing is a ledger your shipping is a ledger everything is a ledger now these ledgers are run by siloed uh participants the public blockchain rails are really this credibly neutral rails that you allow these coordination so how do we kind of make these people easy for people to plug into these ledgers right like those are the real questions in five years I think we'll be able to kind of abstract a lot of this get rid of this tribalistic mental it and you know like bring this technology to real users that's that's what I'm looking forward to we don't have to wait 5 years it's happening of course absolutely well that's time thank you so much guys honor Rock Merck um Nicola thanks for sharing your insights and thanks for joining us today