zkSessions: Mapping the L2 Landscape - zkPorter with Alex Gluchowski (Matter Labs)

Transcript

so this is alex from matter labs and i just want to check what was the name of the talk is about zk porter or what will happen to roll ups if gas prices keep rising take it away great uh thank you anna and uh hi everyone uh my name is alex i'm from mata labs and we're working on zika sync and i'm really excited about this talk because it's gonna be uh to me it's one of these magic moments of zero knowledge proofs where you go beyond things that that seemed possible and they completely turned over your mind so we're going to talk about how roll-ups are going to affect uh east scalability

and uh what are we going to do with with the challenge of gas prices that keep rising uh but a quick intro um interesting thing before i start uh ziki sync is a zika roll up right now it's live on mainnet for payments and we're gonna add swaps uh very soon and currently it's the cheapest uh roll up for payments you can have on mainnet because we we use recursion with planck and we actually go to the limit of what's possible we just closed a an ecosystem funding ground with a bunch of high high profile defy projects wallets and exchanges

and this was done to fund the development of zika sync 2.0 which will bring smart contracts to uh where to on to zika roll-ups on layer 2 that will support composability and evm compatibility so it will be possible to take any existing code written in solidity and probably also viper and transcompile it to our virtual machine and deploy it on this contract with absolute minimum modifications required and it will just work exactly the same way it works on on ethereum so you will be able to call other contracts uh you will have atomic composibility uh everything is going to be

permissionless just like the the exactly the same way you're used to interior and this network is going to be decentralized we're going to distribute the large the like largest portion of the governance rights to the public uh but let's look at yeah by the way we're going to launch this summer so the basic version is going to be working this summer on mainnet okay let's talk about scalability uh so when we started working on rollups two years ago we made the calculations with zika roll up and uh it turned out that we can do approximately 100 x scalability so we could go from 20 transactions

uh 20 payments let's say uh on ethereum with erc20 tokens to 2000 transactions per second which is a visa scale and we thought hey it will probably bring us 100 times more users but what happened was that we just got a lot more trading and uh the gas prices kept rising uh and the users were kind of pushed away from layer one because the transactions of fifty dollars two hundred dollars five hundred dollars for a simple trade are just not sustainable so this is not the the mainstream adoption of ethereum we're hoping for uh and the unfortunately the trend

will likely continue because the trading is just so so locative and all the things around defy make ethereum really attractive everyone wants to be there everyone wants to be to have access to this liquidity but that means that the retail users will face a challenge of uh either you leave the the chain and go to some more centralized cheaper alternatives uh or you just cannot have access to this open and permissionless and inclusive ecosystem which which would be a real pity so uh this is just like what's happening we see this the the number of users grew 10x approximately and and the gas prices have risen 100 attacks

uh like there is really nothing that will stop it from uh from being extrapolated this way even with east 2 um it will say takes some time for for the sharding to arrive so like we will still have this problem for a couple of years so what are the implications uh you want to use roll ups as much as possible especially if security is critical and preferably use zika roll ups because they provide the highest degree of security actually comparable to layer one but we have to recognize that roll-ups are not viable for a lot of use cases and specifically for a lot of user groups like if your users have very small amounts of value and they just want to have the

small transactions or maybe they want to have some some gaming or maybe they they want to trade but they want to trade in really high frequency then roll ups are probably not gonna be sufficient for this for these cases so we need to provide some alternatives and the alternatives we have now on ethereum a naive approach would be to say okay let's give up centralization decentralization sorry let's be more like this centralized layer once and users will trade off security and decentralized properties for for scalability which will already lead to problems

probably with regulation uh it will lead to security problems because there will like the the existing solutions side chains plasma validium i don't know what else they are not offering a balanced security approach so you will have situations where the incentives of the validators will be to actually break the system under certain conditions or the incentives for hackers will be very high to try to attack the system but those are kind of long tail risks which the users might decide to face and take into account but we have a third problem which is really really big uh which is if you go and introduce separate scaling

solutions for different levels of security users are willing to take then you're going to fragment liquidity then all the whales and professional traders are going to be on the platforms where you have a lot of liquidity which are going to be also very expensive and the smaller users are going to be on isolated silos where you have less liquidity like less opportunities for for actually having some reasonable deals so this is going to affect everyone immediately so what we can do well uh i'm gonna uh talk now about ziki porter this is the idea we presented last summer it's based

on the idea of evolution introduced first by ellie bensasson i think it's actually different class so it's hard for me to categorize it's probably false for is probably falling under volition uh but it has very different properties because we're adding decentralized script economic security the way it works is this uh so let's first reiterate how zq roll up works in general you have some state which is kept off chain by the validators and you have a root hash of this state which you publish on ethereum for every ethereum block oh sorry for every block of rollout and you also have to publish through the

call data of ethereum network uh some data for every transaction or for for the entire block of transactions which will allow the external observers to reconstruct the state so that this the the data cannot be withheld by the validators and you keep the security level at layer one so now imagine a system which is just like that with a subtree uh so like you have some of the accounts in the roll-up part they have a separate subtree where all the rules of zika roll up apply so whenever you have an account with an interaction with an account from

this part you will publish all the data so this accounts will enjoy the maximum security and yeah just like you will pay the normal zq roll up prices but in the right subtree we are only going to publish the uh the root hash of the subtree to ethereum and the data the the the like the roll-up data for individual transactions uh is gonna go to the validators and it's gonna be signed by the majority of the validators stake let's say by two-thirds potentially this this parameter is variable what this will mean is these users

will have much less security guarantees because the validators will be able under certain conditions to freeze their their accounts but they will have very cheap transaction costs because they will not need to publish anything on ethereum except for the root hash which is off one just once per block uh and they will still be able to interact seamlessly with the accounts on the rollup site and and the accounts on their own upside will be able to interact with accounts on the azk portal side so what this means in practice is you will have a shared pool of liquidity because you have essentially the same

ecosystem it's just the same single layer too inter-addressable the roll-ups will have layer one security assumptions uh the porter accounts will be under what we call honest minority so if you have at least one third of the stake of honest validators then nothing bad will happen the the users will retain access at all times if you have a dishonest majority they could freeze the funds but this would also mean that they essentially delete their own stake so you will have to burn two thirds of

the stake to burn all the assets of the users which doesn't make sanskrit economically and it also doesn't make sense to really attack such a system from from external uh by external hackers because it's going to be a lot harder to manipulate it so you can manipulate it you can do some black mating or something but like it's it's a lot less straightforward than in the case of uh let's say a uh a side chain or like trying to attack an optimistic roll-up where you can just grab all of the assets from the roll-up or from that side chain and go away with it but the great thing is that the transaction costs in the port our

account part will be fixed they will not depend on the gas price of ether and they will be very very low so the the actual cost of data storage uh we calculated it to be around like fraction of a cent maybe one cent at maximum you will have to pay some additional costs if you interact with the zika roll-up account so imagine you have a uni swap on zika up and all the whales are putting liquidity into it with the guarantees of zika roll up and then you have millions of users who are making trades on this uni swap from

their porter accounts and paying just like fractions of percent so they they still have to share some small part in the in the uh roll up uh state updates but it's going to be very very small so like let's say 0.5 uh yeah so this is great for low value transactions such as micro payments gaming uh some i don't know like something where you you you you really can pay a lot for this transaction you can use it really great for uh high frequency trading well in terms of blockchain high frequency is just like frequent trading we're not talking about hundreds of

thousands of dps but like 10 000 ftps is already possible here and this is great for a lot of users who will be just outpriced by the gas costs even on the roll-ups if things are going to continue on the trajectory where they are now so just imagine next year ether price is going thanks to eap 1559 to like to the moon and the gas price is maybe a thousand gigaway maybe maybe slightly more like let's say two times more than all-time high just 2000 gig away

this is how the cost breakdown is going to look like so l1 is going to be just for like millionaires uh optimistic roll-ups and zika roll-ups are gonna offer some reasonable prices but they're still gonna be in a pretty high segment optimistic roll-ups are gonna be probably like three four three times greater than zika roll ups uh but in the ezekiel porter accounts part you will be able to pay like half a dollar third of the door and so on uh so yeah this is uh i think this is really really interesting uh we can also talk about what this means for developers but i think this is

too specific so i'm only going to talk about this if there are questions thank you very much thank you alex for this presentation um i sorry i was just typing about how brutal those gas prices look i had never seen them articulated that slide before just so high like that's i mean i know you're kind of making a few estimates you're making a few predictions here but uh well you like is 10k for ether realistic i would say it's not unrealistic given all the developments and what we see with uh all the new ways of using things like imagine now only like

imagine all the people from wall street bats join ethereum because they cannot trade there anymore okay and imagine everyone wants to do nfts and and trade and it goes viral and you just absorb really like hundreds of millions of new audiences this is absolutely not unrealistic the demand will surge unless of course i mean the prices are currently holding this demand back right so like the prices will always be determining what maximum amount of users we can handle uh if they are it's like ten dollars then you already eliminate large fraction of users for payments for example all right so like but we we need to plan

for the best case scenario uh if we don't want the users to leave it here totally there's a lot of questions in the chat and i want to bring some of them up um so the first one was from sebastian are there any l2 standards to ensure interoperability from the beginning so there's a question about standards i would say um well we are going to support solidity so like whatever like i think the roll-ups that support solidity have the the standardization out of box so if you if you're if you deploy something on optimistic roll-up uh you will probably be able to migrate to easy key roll-up uh very easily or if you have a state channel

uh that that facilitates liquidity moving between roll-ups the same thing like whatever is possible in ethereum you will be able to deploy it on on optimistic roll-ups that support uh um evm and also on casing uh i i'm not sure about other zika roll-ups but i think that that's very basic and that that that i didn't actually know that so yeah i was i was expecting a different answer there but that's good to hear that there is some sort of standardization are you still there alex yes oh yeah okay are you frozen but i can't see you anymore okay yeah i think the connect when the connections go uh one of us loses video yeah um okay another question

uh zk porter is an interesting approach but how does it help with liquidity fragmentation among l1s and various other rollups oh it doesn't help for that what it helps for is you you can have liquidity inside the same l2 like for example inside the casing and it won't be fragmented between the zika roller part and the prop and the validium part or like the the portal part uh so the because uh like those two parts of the zk porter have different security assumptions so you you like your default option

would be to go and build two different systems with different security assumptions and then you would have to uh to separate liquidity between them but with zika portal you can unite them together and still give people options whoever needs high security goes for a zika rollup and whoever needs lower security because they just can't afford it well they have to trust the validators that divided editors are not suicidal because they they are not going to burn their stake just for the for the sake of uh griefing uh most of the users right so like the the the uh this stake part is very very important here like you you your security assumptions are lower because

people can still do something bad but it's very unrealistic because they have economic incentives all economic incentives not to do that like it's it's completely against their interest cool so we are actually at time for your talk there was a little bit more conversation maybe you can just jump into the chat after this and help to answer it there um and i hope you'll stick around for the panel for the park bench panel and hopefully for gather dot town so we can see you there as well sounds great thank you very cool thank you thank you everyone