The Internet Of Rollups: Presto, zkAppChains And Mass Adoption - Igor Mandrigin

Transcript

uh hi everyone I hope you can hear me quite well uh yeah my name is eager and I'm the C and confounder of Gateway FM before uh coming to founding this company I've been a cev at ethereum for a while at the Aragon team if you heard the previous um uh talk uh there was quite a lot of talk about the stateless client so I worked on those uh for a while now we are also working quite a bit on polygon Z KVM and uh nosis chain we kind of providing a lot of infrastructure towards that and uh today I want actually I want to actually start with a like a tiny demo and then to give

you a little bit of context uh um uh for this demo so I'll just exit my presentation we'll switch here and we'll do one two three clicks and go to the dashboard and what I did right now is is I basically uh started creating my own uh ZK rollup on ethereum Suppa so now we go can go back to the slid and hope that it will actually finish provisioning by the end of my speech uh so we'll see and uh basically why did I do that right now and why did we do this service so um we always were thinking at Gateway about

the good future for crypto we founded this company uh back in early uh 2021 late 2020 and back then if you remember the majority of the ethereum nodes were in AWS in North America and uh we wanted to make an alternative towards that so we make an made an RPC provider that explicitly didn't use public clouds it was reliant on uh private data centers and this data centers was were also located in a European cities and um like they we had couple of of them in Sweden couple of them in Switzerland uh some in

Germany then we added some more and uh that was because like capturing all the nodes like by the same entity like AWS Nam or in the same legislation is uh really really dangerous uh it can kind of give the future that's not really good for crypto when there just a single entity controls everything right and uh right now in 2023 the situation is much much better so we have a plenty of infrastructure providers all over the place from from Singapore to uh Europe to the states of course and uh but now we are going like to the next uh stage of the future and

that's about adoption and I think uh yeah we've been talking about adoption since forever like when I just joined the can ethereum movement in early 2017 everybody was talking about bringing the S bilon like getting Mass adoption and stuff like this but right now it's a it's a bit of a unique uh uh thing when uh we finally have some legislative uh Clarity at least in some uh areas about crypto and a lot of people are thinking about how to how to integrate blockchains and we can kind of think a little bit about uh of those people like in two categories so one is like that I consider myself in is kind of a crypto

native people who are coming from kind of a principal background who understand how the tech works and they are already using the good parts of crypto some ex extra freedoms it provides some extra features some agency and then but uh to be honest this part is kind of in the minority right now and then there is the second big group of people who are interested in the blockchain somehow and those are Enterprises and institutions and theoretically these groups of people can influence much much more people than uh than just a group uh on the left and if we see what's the difference between

these groups are so I would say that it just of course I'm simplifying quite a lot here so don't get offended if you consider yourself in either of this group but like if you're talking about web three natives are we talking about that kind of a this censorship resistance open for everybody permissionless and other principles like this like privacy y y y and if we look at the big companies like any big company and maybe if it's a government agency or something like this uh the first thing they want is to have control because they are afraid to let things go because they afraid to get in trouble and the second things they want is uh to

basically be flexible to um have somebody who can tune these things tune this dials somehow for them again to to feel in control and safe and we want a good future for crypto when as many people uh as possible are using it for like to be more open inclusive neutral without censorship and stuff like this and I think right now we can kind of at a unique phase when uh both um Enterprises and web three natives can finally uh come together in a in a good compromise and I think it's all related

to uh ZKA Technologies uh of course again ZK is the thing that we've been talking to and it was a Hot Topic for quite a while right uh probably a couple of years already it was one of the hottest topic but we only this year we got finally the first serious deployments of this in production so only right now you can come and just show that okay it's not just a proof of concept of a proof that lies at a 5 megab Json file on my computer but there's also the whole system like uh the ZK rollups like of ZK sync of polygon that actually take these proofs verify them use them in the sink use

them in the in the bridges and why I think ZK can power uh something good so to me the future probably will look somewhat like this and the good future will look uh for me would be when we have really good layer ones like nosis chain or ethereum that are spread around the globe well that have like huge vast validator sets that cannot be captured by like a single entity and then we do a lot of kind of a customized layers on top of that uh and that's useful because these layers will still give the sense of kind of a control and flexibility towards Enterprise and and government users for instance but they will still

be uh settled on ethereum or or nosis chain on this kind of neutral Network that are built with principles uh that the the community all always had right and how will thec actually help with that and that's a really good question and I wanted to just tell a little bit about uh about the user agency in ZK rollups and to talk about it basically we need to talk about what is ZK proof what does ZK proof proves in ZK rollup uh if we're talking about just a normal blockchain uh we have the consensus uh that verifies that every transaction it

kind of satisfies a certain set of rules this rules could be that oh your signature is valid the money that you transferred you did have on your account so or the asset that you own is actually your asset and y y there's a lot of rules like this and there's just generic rules of evm and what does the layer one do uh they basically take all the transactions and every node re-executes this transaction and basically they validate this proof by just re-executing because if somebody malicious does something strange uh your noes just will not take this transaction it will reject because it will have another result in uh zk2 the same set of rules is

actually yeah it's uh is being tested and verified by the ZK proofs so there is a ZK proof that generates that all every transaction that since Genesis happened on this layer two always were adhering to this set of rules plus the rules of the evm engine and they generate a proof but the important part here is that this proof is then uh getting verified on an L1 right so it's not L2 generates this proof and it's uh kind of challenges L1 and L1 has actually this script that checks that the proof that you gave to it uh actually adhere to this all of this rules and uh that means that in this

case uh layer one has a big big influence over like a layer 2 operator because uh what happen what's happening on layer one uh basically let's say if I'm a malicious actor and I want to tweak something in the in my rollup then it's mathematically impossible for me to get a proof that will be accepted by the layer one and what's important the verifier of this layer one is very tightly connected on to the bridge and then if you are as a customer like sending transactions or locking your eth into ZK powered bridge that means that if if there is a malicious agent on another side it has way way less power

over you than like in a normal case so that means that uh uh if we have good layer ones uh very distributed uh really nice ones then it means that with these DK Technologies we can bring quite a lot of these properties of user agency to the layer twos regardless of how decentralized they are and again for institutions that want control and stuff they will unlikely be really uh decentralized but there is a problem of course that's uh of course the good future is DK but it's not always that the best tech is the tech that wins uh the market right to to have a chance to

kind of a compete on this Market your Tech should be convenient primarily we see this quite a lot let's say with JavaScript nowadays I don't know if you're all front end developers and will hate hate me right now but I'd say that JavaScript was never designed to build desktop applications from it or to build mobile apps from it but since it was really convenient with a good uh developer documentation with a good Tools around this this one that actually won everything and now every second up on your Mac or on your windows is electron and every second up on your phone is react native but that means that the tech needs to be convenient and also I would say that this Tech it's

actually needs to be built by somebody who cares about principles so because if you can encode those principles very early in the tech and then it's very convenient to use so Enterprise and institutions will use it that means that these principles will be for this Enterprises institutions very hard to change afterwards and that's where my demo actually uh finally uh comes into fruition because what we are building right now and what we're announcing right now and actually launching not just announcing uh at Gateway is uh the thing that we call Presto and it's a first it's a next step into uh what we call the good future uh for um like

blockchain Community basically right now it's a building block when you can create your own ZK roll up in minutes uh which could be public semi-private but they still will settle on ethereum or nosis chain or similar chains and uh we are kind of in unique position because we are also contributing quite a lot towards the km uh implementation so yeah we know what we're doing so let's let's see uh and not only the there will be not only just the rollup of course you you can pick where you put your stuff maybe it's AWS maybe it's a private data center maybe it's some of the public clouds you will get managed rpcs from there you can get custom fee model you

already get a bridge and like a lot a lot of um we are not trying to kind of do it just by ourselves we are uh having on board quite a lot of Partners so these are the partners that we already talked to and discuss there are some because to build an actual useful thing you need to bring a lot of technologes together to build it and might it might be like a kyc it might be some data availability stuff oracles like Redstone uh maybe some nft API or automations like gelato and of course uh something like for account obstruction like NOS is safe uh so basically this is a little bit of how it looks like but let's see

if my rollup actually did provision so yeah it it did so you see it here so I started it about give a take 13 minutes ago and uh now it's live uh rollup on a ZK VM rollup oh sorry sorry ZK VM validum on CA and you can do the same if you go to our website and just well I'll do the worst thing that's our like our team will hate me for I'll try to actually test that it works uh so let's see I have this thing here which is just a valid address and let's see if the faucet oh yeah The

Faucet Works for sure because it's a testest net and you can go to the block Explorer already and you have the first transaction of one is to to the address I sent so it's actually a live L2 uh like provision within like 13 minutes and three clicks and that's what I'm talking about that the tech should be quite convenient so today you can go to Prest to. gateway. FM and just repeat what I did uh and try to get your own uh ZK uh validium running it's going to be free and since it's free you will not have a

real working prover there because for they are too expensive right now to give out for free so if you want a real prover there then you can contact us and we'll provision it uh we also can provision main net so far it's only through contact us forms later this year we are planning to launch uh main Nets as well uh with the bilding that's where you can use your crypto or credit card or invoice uh normal business invoicing uh if you want to see some updates on this uh you can always subscribe to our um Twitter or or X right now for the updates and then you can always yeah as I told you can try it now think you can create one for free uh

of course it's a beta launch it's a test so you might break it as well but that's it so I think that's what I wanted to say and let's leave a little bit more time for questions thank you thank you eigor uh we we have a few minutes for questions I have a question up there uh we have a microphone just give us one second hey h very cool presentation thank you um I was wondering uh obviously polygon ZK evm is really cool

but recently we saw a lot of adoption from op stack like a lot of projects launching there um do you think choosing zkm is a bit like leap frogging beaming us straight into the future or or where do you do you think op St uh sits here I mean uh I think uh op St is a good intermediate step uh but uh what's the big difference between like in general optimistic rollups and ZK rollups and why I think it's important that's uh Z rollups uh has the Z rups provider has to to provide a proof that every transaction on the chain is actually valid according to the rules

like every transaction regardless of what's happening and optimistic rollups they kind of think that's all right everything is okay unless you challenge us somehow of course this challenge could be with the automated as well uh but still I think to me as an engineer it's a much much more neat solution when you actually I am as a potential malicious entity will have to prove to you that I didn't break the rules I think it's much better than you trying to catch me right-handed or some third entity trying to catch me right-handed so we really like ZK rollups and why we choose in particular the polygon zvm is that that

is was the only something that you can call a zkm in production at that point in time and we actually did quite thorough testing by say ethereum Noe ripping off the Merkle tree putting the same Merkle tree that polygon zkm had and running this uh the whole Hermes or polygon ZM main net on like a vanilla evm and then the root hasch has actually M did match so it has a pretty good bite code uh compatibility we didn't find apart from what's written on their Wiki we didn't find any any big differences uh in how bite code is interpreted so it's a pretty cool achievements I saw another hand over

here um yeah thanks maybe to continue a bit on what you just said um a bit of a general question uh can you make a couple of comments on the current state of ZK um rollup technology in terms of how uh how big are the proves how big is how how long does it take to process a proof and uh yeah yeah it's um it's getting better like right now we have something that can be in production that can actually uh prove something like ethereum blocks or layer two blocks every let's say 30 minutes to an hour like everything in them using this recursive proofs and stuff like this

it's still very expensive to run servers that can do proving and uh but let's say right now to run aover and uh we have our head of infra here and he can correct me if I'm wrong but you need to have like at least half a terabyte of ram that's the ABS H 700 okay 700 uh gigabytes of RAM you need like a server for that it used to be that it was over a g over a terabyte so it's getting better and with the new technologies with thing that's at least the guys polygon called variable size State machines it will be even better uh but

so far yeah to to make it so far it's just expensive but it's I think it's good uh yeah any other questions for eigor over here thank you um I believe you mentioned that uh so far the project is active on uh on a test net um I was just curious if you have any stats you could share on projects that showed interest in U using this tool to Shipe their own app specific rollups and uh yeah we have uh a couple like again it was opened uh kind of opened and kind

of first version was launched basically today that means that we don't yet have any stats on on on this version but we with the previous version when we semi manually we're provisioning growups and we have uh two interesting customers one of them is called nosis pay that is currently running in a very kind of old installation on op stack that we provisioned way back in in in the spring right and then we have another it's called Palm network uh which is a kind offt based Network where we are doing the transition of them being this private chain on on hyperledger bezel into a ZK rollup so these are two

clients and we are talken to a couple of more but those I cannot share uh time for one more question if not then let's give it up for eigor thank you so much thank you