zkEVMs: one year in

Transcript

[Music] hi everyone I'm Leia I'm with scroll and I will be giving you a little bit of a service talk today this is a talk for everyone who has been following the zkbm hype it's maybe still a little bit lost but doesn't dare to ask any of the basic questions anymore and I thought the best venue to do that is ECC because this is kind of where it all started the hype around the kabm's last year and um what I'll do I'll do like a little

recap really quickly and we'll also talk about some of the differentiations I have 15 minutes so I won't be able to get into the nitty-gritty details but I think I will be able to give you enough talking points or yeah pointers if you're at a Meetup later today or tomorrow to sound maybe a bit smarter okay how did we get here last year during ECC we had three teams announced there's a KVM plant so scroll we announced our pre-alpha test net the day after polygons the KVM announced that they're going to be the first ones to launch to mainnet which was then shortly after followed by CK sync announcing their 100 days to mainnet countdown if

some of you remember and from that moment on it was really on like the maze the race to minute had started and also I think this was probably around that time when the narrative around L2 Wars kind of was born and that narrative aside um everyone got really excited because everything that we had seen so far um or up until that moment um when it came to the xdk Roll-Ups was um it was not compatible with the evm it was like very specialized and required a lot of special tooling and with the Advent of the ZK VM that changed one year later where are we at we have

definitely a lot more teams working in that space now and they all kind of maybe seemingly do the same really difficult to differentiate um we have three zkavms live now very soon we're gonna have four and with that more teams working on that um we also have a lot more educational material out there there's a lot to read up on but if you're not if your daily job does not involve reading up on ZK avms on a daily basis I think it's safe to say that it's still kind of tricky to follow along and like understand how like how they might differentiate and that's where I'm like now gonna give

a give a little bit of a recap what I think is enough to know um tldr why do we care about the kvms again um we need to scale ethereum if we want to have web 2 like adoption and I know people here in the conference also getting a bit tired of the infra narrative when dabs um but if we want that kind of adoption we need better infrastructure and with ckavms we get the ethereum experience Community deaths smart contracts Etc just in a scalable fashion um then more broadly secure Roll-Ups um we love them because we're they're

based on validity proofs there is unlike for um for fraud proofs There is No Game Theory required it only requires cryptography and math which allows fast finality and fast finality allows the users to do fast exit it facilitates liquidity across layers and all sorts of things that we're going to see more of I'm sure in the next month's years also not to forget we do love when things trip early because in the past we we have not been shipping things to quickly or too early and this was mainly possible through better proof systems

Hardware acceleration but also um things like the PSE team like a large large-scale open source circuit engineering and yeah leading me so we established why we love UK avms um but with like also after ACC last year more and more teams coming on [Music] um yeah it was it was difficult to like understand okay what what is everyone doing and in August that year a vitalik published this distinction or classification of the different CK EVMS and I think that it's important to

understand the evm was is not built in a very seeky friendly way so if you want to match the ethereum experience you as someone you have to make some trade-offs and there's a team working on this you have to make different design choices so if you want to be asked like Max compatible you're gonna you you're gonna have to compromise on performance for example and so so that's why we have this classification now and it's very widely and commonly used I get questions about that um every time um I do I think it's perfect maybe um it does help but it's also very

technical so um I'm nonetheless I'm gonna run through it like really quickly in like one sentence each so um you have the two axes you have compatibility and you have performance and type one is um basically ZK CK ethereum as if ethereum wasn't a snark then you have Type 4 which is on the other side of the Spectrum which is a ZK friendly VM that is not compatible with the evm and in between you have three more distinctions that are Technical and are a little bit more subtle type 2 is basically evm equivalence but not ethereum equivalent so you have the same smart contracts but you have seven

changes to the protocol 2.5 does have some evm modifications such as gas pricing for example maybe you've had that on crypto Twitter floating around um and then type 3 is more evm modifications um like for example pre-complies Etc and most of the teams out there are probably working towards like two or two point five and the ones that are currently in production are probably more on like three four and um as I said while I do think it's it's helpful um it sometimes also really tricky to to learn about these subtle differences

so what I find a little bit more helpful to step back and go ahead a bit more high level and basically having like these two these two buckets because it focuses a bit more on the impact rather than these subtleties and so language level compatibility is also a mouthful and byproot level compatibility language level compatibility you compile an evm friendly language to a snark friendly VM that differs from the evm but there this comes with potentially compatibility issues and but also what you might want to consider if you're looking at like where to deploy your dab it might also mean that it you can move faster on like

some experimental apis for example or include like crazy op codes so maybe you're gonna resonate with that one you also have then if you then look at the bytecode level compatibility side there you can just interpret the evm byte code directly and it makes it just really easy to Port over your dab if you have deployed on on ethereum on the evm Chain already already uh because it's compatible with ECM smart contracts currently we're looking at that it is potentially slower in performance but the teams working on this are continuously optimizing and um think that we will sort of like close that Gap more and more and gradually and

is also closer to ecm's roadmap so the teams can also potentially contribute to ethereum's roadmap upstream and this is also where you can put scroll into that bucket um so looking at that and I think the talks that you've seen before me um already hint at that or already gone to that because there is a lot more to um to a roll up than the evm there is um so if you look at the different teams you might also want to look at and I think this will become more and more of a decisive factor in the months to come

you might want to look at the General in like General Innovation like what does the team think about what comes next for the L2 how does it think about like what use cases does the bet on Etc then very fitting with the previous talk about the sequencing part so um the what is really important also for a team is to remove the single points of failure um and how the teams think about decentralizing their sequencer Network um open source spirit I think this is why we're all here um teams approach this in a different manner

um how how transparent our teams how much do they also collaborate with different teams there's like a ton of um there's a ton of design spaces that still needs to be figured out and there we're also going to rely on collaborating with different teams security is a no-brainer but is this is going to be just like really important for the teams for for any team going forward and most likely no team will be successful if they don't take that seriously in the long term so that includes audits but as well L2 Mev or also the multi-prover setups this is obviously again um an incomplete list

um if you want to or if you want to continue like tracking how the different teams are doing on that um l2beads is a really good one and just to follow along they list some of those aspects on their on their website for the chains that are live and one that I recently found um which I also think is is interesting to take a look at is roll-up codes it goes more into like the technical um differences of the different l2s in general not just ZK roll-ups and um with that I also want to obviously want to make the case for scroll um

you can as I said earlier you can put this in the by compatible compatibility bucket and the feedback that we've been consistently been getting from developers at hackathons and hacker houses is that it is um it is almost boring to deploy on scroll because um the the developer experience is very close to ethereum L1 and you don't necessarily notice that you deploy an L2 and all you have to do is literally change the RPC um during and then also like one thing that is really important for us as well and during East Denver we had like all these memes about

who's going to be the first zika evm who's going to be more first than the other um and around that time we were thinking about maybe it's equally important to think about who is Community First or like or to be Community First and what we mean by that is that we think it's really important to build out in the open that's something that we have been doing from day one together with the PSE team and like the theorem foundation and I think that also should trickle down in how you engage with your ecosystem so we have last week we launched our initiative contribute to scroll where we want to hear from Individual

contributors but also different teams what does it take to um to get them to the next step and but also how can they contribute in like shaping the future um of scroll Security First at every step you're not gonna see us rushing to Maynard without a proper audit we are taking precautions there l2s are something you built for the long term that's what we're here for and I mentioned it before and decentralizing the sequencer network is going to be important for every L2 this is something that we have been thinking about also from the start we are

exploring different um models in parallel and we will shift our our Focus towards this um right after we launched mainnet and which brings me to my last point mainnet we are going to provide the ethereum ecosystem with the developer experience that you've had that you already have on testnet very very soon so we're going to have 4z kvms we're gonna launch end of um end of the summer in Q3 and oh I'm really good on time um and then yeah just to like wrap up I think this week is busy for everyone

here um we're going to be at a bunch of different events we have a lot of people from the team here we will be speaking discussing would love to talk to everyone this is a quick overview and up next um one one event that I want to highlight is that our team is very very distributed we have a lot of people also in China and we're gonna host the very first eth um yeah eighth event or um in Hong Kong um in October this year and obviously also deaf connect this year we're gonna host an old today together with l2b which will be exciting if you want to

participate come talk to me afterwards and I think that was it that's all I wanted to say thank you okay we're I think we're headed into lunch so I will be here if you want to talk to me have questions we can also just discuss outside cool thank you thank you