Rollup Wars / Will Harborne

Transcript

hello everyone so I'm will oh I'm the founder of rhino Phi which is an all-in-one platform for defy uh we make it safe and easy to access D5 on every different chain fully self-custodially and I'm going to be speaking about the number of Roll-Ups we now have uh talk about how we got to this position um then look at some of the statistics to how Roll-Ups are doing how the competition is going between them and some of the tools that I think are going to be really important for how Roll-Ups differentiate against each other uh and how they are gonna sort of combat each other to try and give us the best user experience and then what piece could look like eventually in a few years time in the roll up space so let's start off with how did we get

to this position where there are so many Roll-Ups out there um and actually it's not just Roll-Ups we've had this evm Cambrian explosion where there are just so many evm chains and in a lot of ways they look very similar and for users this is incredibly confusing um and but you know in a way this is this has come out of a result of a demand uh for Block space so in the beginning we had ethereum um and ethereum was great we could uh build smart contracts for the first time and we saw the emergence of D5 and we thought that was going to last forever but actually very quickly as as defy and crypto usage grew uh we had congestion we had high gas prices and we had slow

block times and so back a couple years ago now four or five years ago we had this idea that layer 2 would just come along and solve all of our problems and these were some of the very early layer twos that were being talked about maybe three years ago um and some of those of course are still around today and still have become what we're talking about for Layer Two but the reality is what we have looks a lot more like this so there's a whole bunch of layer twos or things that call themselves layer twos there's also lots of layer ones which are evm compatible there are things that call themselves there too is but actually they're ones there are maybe layer threes soon on the horizon and there were all these complex Bridges connecting them together with different risk assumptions and trusts assumptions

um and it's a mess and there's there's already a competition starting to form over why Builders or users would go to these different chains but it's very difficult from a user perspective to figure out how to do this safely which chains and EVMS you can trust and which are uh which can be used and so we need to get out of this situation we need to find a solution for people to really trust defy and crypto and be able to safely use this layer 2 Network and we want to make it feel like one seamless experience and that's just the beginning I think over the next two years we're going to see thousands of Roll-Ups launch it's going to get even easier to Launch a roll up but 95 of them will never reach sustainable usage and we saw the same

thing happen in the layer 1 space over the last few years and also in the most layer ones never got adoption never got used um but there was a lot of energy that went into building those and using them and so we're going to see this technological race and a really brutal competition for attention and and a price will potentially play out which in some ways may be good for users because we'll end up with um a great set of secure layer twos at the end of it but in the meantime may be bad for this ecosystem and make it harder for users to navigate so let's look at what is actually happening right now in terms of uh the way that Roll-Ups have evolved and uh the statistics on them being used

but before just before we do that I think we should consider what is the business model of a roll-up or a layer two and really these are two-sided markets like any blockchain they need to build an ecosystem of applications and liquidity that people want to use and then they need to also attract the users to come and use that ecosystem if they can do that then they can charge transaction fees um just like any any chain and then they pay a subset of those transaction fees to the layer one for the security that they're essentially renting from for example ethereum or in future it could be from things like Celestia for data availability or other layer ones obviously if the amount of fees that they're able to charge to end users is higher than the security they're paying

then it's a sustainable Layer Two um and they've reached kind of that sort of critical mass so if we look today um there are some Roll-Ups that have already reached that sustainable level of fees um it could be compared to what they're paying so arbitrim is an example uh that has really seen big revenues now actually they are charging they have enough users and enough critical mass to be charging transaction fees and actually they have a 30 to 40 profit margin currently um on what they're charging versus what they're paying to ethereum and that gives them quite a high annualized Revenue which they can put into the Dow to continue to invest into building that ecosystem um and optimism is in a similar position

still growing very fast but most other Roll-Ups have not reached that point yet and most probably never will um so there are there are several levers they're going to have to think about in terms of going head-to-head um first of all those two points about building an ecosystem and then attracting users to it and then second what fees can they really charge and what are their costs going to be for the security and the competitive pressures that we're starting to see play out now as we get more um more and more Roll-Ups is going to be good at pushing all of those to really be optimized and making people have to fight over the attention of of us as users of these roll ups

if we look though over the last few years and see where we are now so the top line is ethereum's market share of tvl and it has gradually dropped uh and then we've seen correspondingly some growth in most of these other evm environments or layer twos you can see a small green line down at the bottom that's arbitrim so on the grand scale of things really the the sort of market share of most Roll-Ups is still tiny and arbitrim is in the lead um and so this is not a game that's like yet played out although often when we're looking at Roll-Ups we think that you know maybe the biggest ones have already emerged there's still a lot to to play for and um if we if we look again at that those stats

arbitrine for example has grown from 1.8 to about five percent of tvl in D5 but that's still very small numbers um and most of the others are still nowhere near that and for example Tron and BNB chain a larger um and and so we it's really very early in this competition if we look at tvl arbitrim optimism and ZK sync are kind of the The Front Runners um in terms of open evm environment roll ups um arbitrim at about six billion dollars tvl but most way below that um and same with transactions per second actually we're getting closer to what is going through ethereum for some of these um but the numbers are relatively low

compared to what they're set up to be able to handle um there are some sort of honorable mentions like starknet uh which is which is growing steadily but is a bit earlier in that life cycle and also polygons EK VM which launched very recently but growing quite fast but basically there's still hundreds more to launch so even in the next few months we've got Base by coinbase Linear by consensus scroll coming later this year and a stark net sort of full launch so it's it's gonna get a lot there's gonna be a lot more competitive pressure um and so we need to think about how that's going to play out and what the experience is going to be for users trying to navigate through that and on top of that there's things like Celestia wool kit um which stays literally is its goal to

make it possible to deploy a sovereign roll-up on Celeste in just minutes um so we're just at the beginning here the actual infrastructure of launching a roll up is going to get so simple that literally anyone can do can do that and you can imagine for example launching your own roll up for a product or an application where you have a bit more control and able to fork and upgrade it um compared with launching on arbitrary or optimism or some other public rollup where you don't have that that power as a as an application Builder so if we're going to see all of this what sort of tools do these Roll-Ups have to really um encourage them to become the largest uh ecosystems and to attract Builders

and users and I think in the short term the main factor is still cost um so as rhino5 we support six different um Roll-Ups as well as a few other evm chains and what we already see is that cost is a big sensitivity factor for current users of Roll-Ups um and this is this data here was taken on quite a low uh cost stay a few days ago ethereum around 4.5 dollars for a simple transaction um whereas polygons eke VM and ZK sync the the kind of ZK Roll-Ups a bit more expensive than optimistic Roll-Ups um but not much in it uh and we're talking in the kind of 10 cents to a dollar range almost all

the time when gas prices get high that can still get uh still get important for users um and there were quite a few levers already available today to to sort of change those and allow these um Roll-Ups to compete so ZK Roll-Ups uh are currently more expensive most of the time but they have an advantage at higher transaction throughput so as they get more and more adoption their overall cost per transaction goes down because uh of the way that the the Roll-Ups amortize the cost um whereas optimistic Roll-Ups don't necessarily benefit from that quite in the same way and all Roll-Ups are soon going to benefit from eip4844 and projecting which will reduce the cost uh for

putting data on chain and mean that uh basically all these numbers should come down which will give them a choice as to whether they capture more of that as basically Revenue through their sequences or whether they pass it back to users in order to compete and attract more Builders and users and there's also going to be choices around data availability but long term what we can expect is that this probably won't be the the comp the competitive ground because the cost should come down everywhere Community is the other area that I think we're going to see very much in the short to medium term a lot of competition and this is really about ecosystem building is hard we've seen how hard it is from layer ones uh where many chains with incredible technology

didn't manage to get critical mass and get you get Builders to come there once you've got a head start as we're already seeing happening with places like arbitrum and optimism maybe it's it's very difficult to convince people to take a big risk on some other roll-up and so it's going to get even harder if you want to do this you need to have really strong grants programs you need to build a real strong sense of culture and excitement in the way that ethereum did in the early days um and we can see good examples already out there for example ZK sync launched very recently and did put a huge amount of energy into ecosystem building before launching and therefore has managed to quickly jump into one of the top few chains but this is something you need a lot of you need a big team a big amount

of funding in order to do this properly it's not going to be a case of best technology wins um I I don't think token economics is the other big lever that ties into that airdrop hunting airdrop incentives has been massive in terms of getting these first few Roll-Ups live and I think that will continue to be the case but there has to be then the ability to build a sustainable ecosystems to then capture that and keep people coming back which we haven't necessarily seen seen evidence of yet remember even those leading Roll-Ups like arbitrary and optimism are quite small in tvl compared to ethereum for example and so that may not be enough uh to say for sure that they have uh really something which can't be take can't be stolen away and

we saw for example um on the on the chart before polygon and some other evm environments that were layer ones have lost a lot of market share recently so even what they gained over a year ago didn't necessarily stick around always um and then interoperability is the other one so one thing that these Roll-Ups can compete on is how fast finality is and that is basically when blocks are finalized in these Roll-Ups and the shorter the finality time the easier it is to bridge to these chains um ZK Roll-Ups can theoretically have very short finality times um whereas optimistic rollups need to wait a little longer to verify things but there's all sorts of things that are now blurring the lines here and that's a

big area of competition between Roll-Ups um so you can have for example committees that give give soft finality by basically verifying that blocks have been seen in the sequencer and won't be rolled back and you can also have Hybrid models between optimistic and ZK roll up which may increase costs slightly but give shorter finality um and from what we're seeing at the moment in terms of adoption having fast low-cost bridging is the single biggest driver of growth uh in user activity and I think long term will be one of the things that can really give um very strong competitive advantages to specific Roll-Ups or evm chains and quality matters more than quantity on

these Bridges because if you can move quickly your funds on you're more willing to experiment try it out and same with moving them off and just anecdotally at the moment the trend tends to be that the the chains that have been sort of most successful in gaining tvl and getting new users to come and at least experiment and try them out have been the ones with the most Bridge the most Bridge options so let's think about as these factors all play out over the next few years what could peace eventually look like for the roll-up ecosystem what do we want to see happen and one thing that we could see happen and some people think we will see would

be a single Survivor a single roll-up that becomes dominant that everyone comes to use and where you can get basically the same experience we used to get on ethereum when it had a small amount of users where you get full composability and you don't have to think about all these different roll-up options now there are some reasons why we don't want to see that if we think back to the kind of business model of a of a roll up um if you're charging transact if if a single roll-up ends up being the one that exists they basically have a monopoly they can charge whatever transaction fees they want and just pay a small amount of that as security to uh ethereum and so I think for all of us we should want to we should want to avoid a

monopoly over this um although there are ways you can mitigate that with for example decentralizing sequences and other things I think more likely we want to see many different Roll-Ups with different trade-offs um living in harmony and if we want that to happen there are a few things that we need to work on now as a community to make sure that we actually get a decentralized ecosystem of different Roll-Ups with different um characteristics that feel to your users as a seamless experience first of all we want low risk protocol level interoperability something like IBC in the cosmos ecosystem and second of all we want more shared infrastructure so for example shared

proofers shared data availability shared sequences and there are teams now working on these sorts of things but they're not being used yet in our existing world-up ecosystem and then differentiated cultures and community so clear reasons and clear trade-offs that have been chosen by different Roll-Ups as to why you might build one application on one roll-up which might be cheaper but less secure and another on a different one which is very high level of decentralization but more expensive um and I think it's important to think that users themselves are very likely to be infrastructure agnostic they won't care which chain or layer 2 they're on as long as it serves their needs and they can move quickly between them but applications and products will maybe

care and will will want to be able to have different levels of trade-offs on security cost and sovereignty so at one if I will but we're trying to be an aggregator layer for all Roll-Ups uh we want it to be a seamless experience to move between these different um defect ecosystems and for the user not to have to worry so much which one they're on so already you can go through most of the top roll ups um through rhino-fy optimism arbitrim Etc um and you can bridge between them you can also go into yield opportunities on any of them self-custodially uh with basically one click of a button without having to worry about whether that opportunity was on arbitrim or polygon and know that it's been sort of vetted

and aggregated into this uh one sort of user experience tracking your funds everywhere um we think that's really important but now because it gives this feeling that all of this ecosystem is a single seamless ethereum extension rather than many different blockchains competing with each other um so we have support for Stark X and actually the core infrastructure that we use to enable all of this uh is Stark x uh built by the starkware team uh way back before they were focused on starknet and that gives us the ability to give a really uh consistent user experience across all the other chains that we then support for example we have uh this bridge option which is one of the most popular

things it's on winify uh we've had over 300 000 transactions through it you can basically bridge between these different um uh chains in 15 seconds um so a few takeaways that I kind of just wanted to go back to first of all in the next 12 months is going to be possible to launch a roll up in minutes but just because we can doesn't mean we always should launch more of these Roll-Ups I think if we don't address the intro interoperability between them we're going to see this war get even worse over the next few years as there's just more and more competition for attention over very similar looking roll apps and for users that's going to be a very bad experience

um and so instead we should think about the fact that users won't care necessarily which layer tree they're on they'll just want it to be a seamless experience and I do encourage you to try out rhino5 if you want to have an idea of what that uh sort of layer 2 D5 future can maybe look like at an early stage thank you very much [Applause] any questions um does it work amazing um so your main page was um it and I agree with you it will be very easy to you know bring a role up on

ethereum and uh your main idea was it's not should be like that that a rollups needs to be like there will be one winner look there will be more and it should be like this according to ethereum values but the thing is like you're bringing like L3 and how you will um how you will define whether to take the project or not since it will be really easy to launch a roll up you as a project need to you know somehow sort if we take this project or not and it will be not also like a monopoly point if I understand this I am correctly what is like how you will solve this problem personally I think there's different schools of thought on this but personally I think that they kind of

user layer interfaces have a responsibility to do betting in this space and to figure out what is secure to look at the different trust assumptions of different Roll-Ups or projects to look at Audits and to sort of curate a little bit but of course there'll be multiple you know platforms doing that that then mean that we still have access to everything in in defy I don't think it's enough to just say oh uh you know it's an open ecosystem let anything go on here because you know and we've already seen this there are there are plenty of evm chains which call themselves for example layer twos Roll-Ups but which aren't or which have very centralized elements to them to their Bridges or whatever else it might

be um and I think as D5 we need to be working on trusting credibility at this point if we want to see real adoption and for D5 to be a genuine alternative to cd5 or C5 um so yeah I think I think we have a responsibility to to curate and to look at what's safe and um yeah how does it work right now how you onboard the project how who decides who is there um within like if we will onboard the project or not and what is a long-term vision of the your protocol at the moment that's that's a decision made by our team and I think it's it's still relatively low effort to do it uh just because there's smaller options but

I think in future if we want to be transparent about it they need to be very clear kind of guidance and and criteria that maybe can involve more kind of users and Community yeah right so now currently you're like you're making decisions within your team centralized one do you have like in a long-term Vision to decentralize this um thing because like you understand it's a really critical point because like you you're bringing ethereum and what you're doing like like you're solving the problem of centralization decent centralization without with with an argument it should not be like one L2 and what's your start of it doing like uh you know like it's an I I also think

like a lot of people have this question like this is I just asked this I think it's a good point we haven't really in detail discussed that but I think it will be important to have very clear open criteria I don't think for example user voting is the right way to go about that because ultimately I think it's the the safety that's important um but I think having very clear transparent open processes will be probably the direction we'd go in hi thank you for talk um so uh could you uh shed some light maybe if you put some thought into that uh what do you see the evolution of privacy Tech and in particular is your knowledge for privacy in the roll-up uh Evolution for the years to come because one thing that

makes me slightly concerned is that we have this point of centralization which are the roll-up sequencers uh and users transactions are not just available for anyone like on layer one but they are also aggregated at some Central points so it would be nice to see uh in my opinion some fancy cryptographic tricks to preserve users privacy and how do you how do you see that Evolution yeah I think there are already some really fantastic teams working on that problem and you have ZK ZK Roll-Ups uh Aztec for example did launch a really fantastic uh sort of uh stage two of their their Vision uh which allowed you to access ethereum defy in a private way but they've now closed that down and focusing on the next stage of their

Vision um and I think ultimately it's true that if we want to see very large scale adoption and personally I want D5 to be an alternative to the traditional Financial ecosystem it will need to be private I think that's a basic requirement um but the main reason why that hasn't I think got big adoption yet is the cost um and it's you know it's more difficult to do and it's more expensive to do so it at scale I think those um will be maybe the the winners but uh for now I think there's too many other problems to solve around interoperability and other aspects I think we have one more question uh thanks Bill and congratulations for

the presentation really nice and on point uh my question is uh do you or are you working with an interoperability a generic interoperability uh partner for achieving your vision and number two would an interoperability protocol with a composable middleware which enables logic based bridging or uh you know with ability to track state changes between layer ones and layer twos or between layer twos would it be helpful what's your views on that yeah great I think that's a great question um and I think the the state of current interoperability protocols uh especially across across layer twos is still quite

early um there are great sort of options out there from a technology perspective um I mean I won't name many of them but you know seller hop for example um but the main problem I think at the moment is most of those uh sort of at the moment have optimized for very high level of uh decentralization require high class realization so they're expensive to use and slow um and I think longer term what we would hope to see would be many of these protocols becoming more baked in at the protocol level of Roll-Ups um so one example I like is polymer which is focused on doing basically bringing IBC between optimistic rollups and I think personally IBC which I

mentioned before uh which is in the cosmos ecosystem is kind of the sort of end State we should be aiming for between Roll-Ups so la like client Bridges between Roll-Ups with lower levels of trust assumptions maybe lower collateralization requirements um but we're not there yet so for Rhino fire we've sort of made the short-term decision to wait and see uh before too heavily betting on any one protocol for example so we're keeping very relatively agnostic and waiting to see how that user experience will improve um yeah cool thank you very much [Applause] [Music]

foreign [Music]