EthCC #4 | "The Polygon zkEVM Frontier" - Mihailo Bjelic

🇫🇷 Welcome to the Bankless EthCC 2022 Experience 🇫🇷 With fantastic guests from all corners of the ecosystem, this 8-part series is an exploration of crypto culture and the current state of Ethereum. In episode 4, Mihailo Bjelic joins us to discuss the announcement of Polygon’s zkEVM. This explosive headline comes alongside three other zkEVM announcements. Why now? What does this all mean?! ------ 📣Rhino. Fi | Massive Mystery Airdrop https://bit.ly/3o9trRE ------ 🚀 SUBSCRIBE TO NEWSLETTER: https://newsletter.banklesshq.com/ 🎙️ SUBSCRIBE TO PODCAST: http://podcast.banklesshq.com/ ------ BANKLESS SPONSOR TOOLS: 🌱 LENS | ACCESS CODE: WAGMI https://bankless.cc/Lens 🚀 ROCKET POOL | ETH STAKING https://bankless.cc/RocketPool ⚖️ ARBITRUM | SCALING ETHEREUM https://bankless.cc/Arbitrum 🦁 BRAVE | THE BROWSER NATIVE WALLET https://bankless.cc/Brave 🌉 JUNO | BRIDGE FIAT TO LAYER 2 https://bankless.cc/Juno ⚡️ ZKSYNC | THE LAYER 2 SCALING ENDGAME https://bankless.cc/zkSync ------ Topics Covered: 0:00 Intro 3:00 Polygon zkEVM Announcements 6:50 How it Works 10:00 Road to Maine 15:30 EVM Equivalence 19:30 EthCC 2021 ------ Resources: Mihailo on Twitter: https://twitter.com/MihailoBjelic Polygon: https://twitter.com/0xPolygon ----- Not financial or tax advice. This channel is strictly educational and is not investment advice or a solicitation to buy or sell any assets or to make any financial decisions. This video is not tax advice. Talk to your accountant. Do your own research.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:03David Hoffman?

  Welcome back to the station, back to the ETH CC experience. We are almost halfway through. This one is with me, Mihailo from Polygon. Polygon was super hyping up their announcement, which they announced at the ETH CC number six conference, which was their ZK EVM. And as it turns out, fate would have it decide that they were actually just one of three teams that decided to release their ZK EVM during ETH CC week. ECC is big like that in the sense that people wait to release their announcements for ECC, or they like rush to get their thing announced at ECC. ECC is actually, you know, it's actually quite a big deal as far as conferences go. And Polygon waited for ECC to release their ZK EVM, but the ZK EVM wars are on.

• 0:48David Hoffman?

  So we talked to Mihailo about everything they're doing with the ZK EVM at Polygon, and of course, all the other peripheral topics around Polygon. And there's a fun little backstory, because I was talking to, hanging out with Mihailo at the last ECC, and which is really where this ZK EVM announcement really got inceptioned behind the scenes way back when with a phone call, which I was just like sitting with my good friend Mariano Conti, having a beer with Astrid and Mihailo when he received a phone call. Fast forward one year later, they release a ZK EVM testnet as a fallout of that phone call. It's just a fun little story. So we tell this story and more in this fantastic episode of the ECC experience. But usually, we do our full pre-recorded ads, but instead, I'm going to do them all right

• 1:32David Hoffman?

  here, right now, super fast, so here we go. One of the big themes of ETH CC week was the launch of the ZK EVM. Three different teams all launched their testnet. And while everyone is trying to claim who was first, we can confidently say that ZK Sync has been focused on the ZK EVM for the longest. The golden endgame of the ZK EVM has been the vision of ZK Sync since the beginning. And during ETH CC week, they launched their 100 days to mainnet plan, which is now already down to 93 days. So if you're going to join the ZK EVM wars, make sure you post about it on Lens, the first decentralized social networking graph. In crypto, we say not your keys, not your crypto, but on Lens, we say not your keys, not your content. Own your own social media handle, choose your own algorithm, and mint your profile using the secret code word in the show notes.

• 2:16David Hoffman?

  And something that's likely going to be talked about on Lens is of course the Ethereum merge. And the merge is why you should be staking your ETH with Rocketpool. It's a decentralized Ethereum validator network where you deposit ETH and get your rETH in return. And then you can take that rETH into DeFi. And if you run a node, you can let other people deposit their ETH into your node, and you earn 15% of their staking commission as payment for your node operating services. And then maybe you want to take your rETH over to Arbitrum. For all of your DeFi things, Arbitrum is where you have to be. With over 35,000 contracts deployed and 1 million unique addresses, Arbitrum is leading the way into the age of rollups. You know how I know those numbers? Because I watched Arbitrum's talk at ECC, which you should definitely also watch, but not before you're done watching this video first. But in order to get your money over to Arbitrum so you can buy your rETH natively,

• 3:01David Hoffman?

  then you should use Juno. Because Juno is the fastest bridge from your checking account to an Ethereum Layer 2. A checking account to an Ethereum Layer 2 is absolutely crazy, and it just takes minutes. Juno is your crypto checking account for the crypto era. But when you're doing all of these things, make sure that it's on a privacy-first browser, which is why you should be using Brave. It blocks all the browser ads, it's got a native Web3 wallet, and it puts the user first the Web3 way. And now I bring you my conversation in person with Mihailo from Polygon. What's up, Mihailo? How's it going?

• 3:33Speaker 01

  Hello, David. Just all over the place with all the great announcements coming from the Polygon side. And yeah, thanks for taking your time to chat.

• 3:42David Hoffman?

  Oh, of course, of course. Every time I see you at a conference, you're always moving a mile a minute. And it's always because you have something to announce.

• 3:49Speaker 01

  Yes, fortunately, it's like that, yeah.

• 3:51David Hoffman?

  Yeah, it's a good place to be in. So what is it this time, Mihailo? What did you guys announce this time?

• 3:55Speaker 01

  We are very, very excited. Today is a very big day for Polygon. I would say it's probably the single most important event ever since we announced Polygon. Today, as you probably know, and some people already saw the news, we announced ZK EVM. So Polygon is the first project that actually shipped ZK EVM, which is basically considered the holy grail of scaling. So ZK EVM, for those who don't know, is basically a ZK-friendly EVM implementation. And EVM is, as most of us know, I guess Ethereum virtual machine, the core execution component of the Ethereum itself. And for quite some time, we have been, as a community, pursuing this holy grail of scaling that is going to have this free... Major, very important features in one.

• 4:40Speaker 01

  One is, of course, Ethereum compatibility or ideally VM equivalence, which is even a step further. We ideally want, we all know Ethereum is absolutely dominant in terms of developer mindshare, in terms of tooling, the maturity of the ecosystem and everything. And we, of course, want to leverage that. We want to stand on the shoulders of the giants, of the giant, and we don't want to waste all that great amount of work that the community has put so far in Ethereum and building all these tools and bullets and everything that we currently use. So with the VM equivalence, we get to leverage all that and keep the same, exact same developer experience that developers are used to, same programming languages, same developer environments, et cetera.

• 5:25Speaker 01

  So that's what ideally we want. That's the first major feature. The second major feature is, of course, security. So we know Ethereum is the most secure programmable blockchain in the world, and ideally, of course, we want to leverage that security because in Layer 2s, we still have millions or hundreds of millions of dollars, and we really want, we really need the highest level of security, and that is what Ethereum gives us. So ideally, we want to leverage that. And the third major feature is scaling, of course. We need to scale significantly. We need to scale orders of magnitude if we really want mass adoption and if you want to onboard billion-plus people to Ethereum. So this was that holy grail that ZKAVM has been representing until today, basically.

• 6:09Speaker 01

  And we are very fortunate and incredibly humbled that we were able to actually ship it. It's full-featured, it's fully open source, which is very, very important. Anyone can take a look at the code, examine, make sure, first of all, that it's really a full-blown VM implementation with all the important top codes and everything. And then, of course, we want this to be a community effort. And like everything that Polygon has done so far, everything is, as I said, fully open source. We invite community to take a look at what we have built in the background throughout this past year to give us feedbacks. We expect contributions from other community members. So yes, all in all, very, very exciting, very exciting moment.

• 6:48David Hoffman?

  Yeah, okay. So ZK and VAVM, like you said, is considered like, there are no further, like, scaling. Like degrees beyond a ZK EVM. And since this is a developer conversation, this is going to take me a few questions to like really fully unpack this. So, you know, everyone, there's general consensus that a ZK rollup is like the end state, kind of as we said. It's that EVM part that's the difficult part. We have like application-specific ZK rollups so far, like DYDX chain used to be this. We had like Diversify used to be like a Stark X like app-specific chain. A few others, like Immutable X used to be a Stark X app-specific EVM chain. It's that generalizability, which is the ZK EVM, so the Ethereum virtual machine

• 7:33David Hoffman?

  inside of a zero-knowledge rollup. And so like that's what we're talking about. So that's been like the end goal. So what was actually shipped today? Like a mainnet? Or what was actually shipped?

• 7:45Speaker 01

  There is a road to mainnet that we today announced. Of course, it's a very, any sort of mainnet is very high responsibility, especially given that we're really seeing huge level of interest for ZK EVM. Everyone would like to try it, everyone would like to deploy, people will transfer capital. So we want, today we announced this road to mainnet. Today we announced the code. We shared it with the world. The testnet, internal testnet is basically days away, probably a week or something like that. Fully permissionless public testnet is weeks away, probably one month, roughly, one month, month and a half. It depends on several factors, but it's really, it will be live in a matter of weeks. And after that, the mainnet.

• 8:27Speaker 01

  Between public testnet and the mainnet, we really want public scrutiny, we really want multiple audits, we really want battle-tested, you know, testnet that has been proven quite a while. So it will definitely take several months, two to three months, I think. So we're basically targeting end of the year for the final mainnet release. But what has been released today is the code itself, and everyone can actually see that that we have basically re-implemented the whole EVM, which was, as you said, exactly the hard nut, the tough nut to crack, basically. Proving some of these expensive op-codes of Ethereum like Keck, ECDSA, and some others

• 9:08Speaker 01

  was really, really hard, to the point that some people were questioning, is it even practically possible? We were very fortunate that These great teams that we have under the Polygon umbrella, Polygon Hermes, Polygon Zero, and Polygon Miden, there has been several cryptographic and engineering breakthroughs, very major ones. For example, on the Zero side, Polygon Zero side introduced this Plonki2, which was the fastest recursive prover in the industry, basically. Goldilocks Field, which is this very specific arithmetic field that has some very interesting properties and allows us to get some very significant efficiency improvements. So combining these breakthroughs

• 9:48Speaker 01

  and really great collaboration between these teams for the past 12 months, we have been able to crack this hard to implement top codes. And yes, we are incredibly fortunate and excited that

• 10:04David Hoffman?

  this happened. Okay, so what happened today was the starting pistol for the road to mainnet.

• 10:10Speaker 01

  Yes.

• 10:10David Hoffman?

  So you guys published your code for like public peer review. People get to look under the hood. It's like it's finished code.

• 10:16Speaker 01

  Finished.

• 10:16David Hoffman?

  And so like, and then soon there will be like-

• 10:19Speaker 01

  Including the prover, which is the hardest part, the whole implementation and the prover, everything is open source, everything, and it's full feature.

• 10:25David Hoffman?

  Right. Okay. So you've also published this and like other people can take this code and run with it if they so choose.

• 10:31Speaker 01

  In a sense that anyone can obviously use the code, examine the code. On the license itself, we are still deciding. It's definitely going to be an open source license. As I said, we open source the code today. We really, and we definitely make it available to everyone to reuse moving forward. We're just deciding when the exact open source license format that we will have.

• 10:55David Hoffman?

  Beautiful.

• 10:55Speaker 01

  But it's definitely the, it's definitely going to be the contribution to the community that people will eventually have, will be able to use.

• 11:04David Hoffman?

  Okay, so code is out there, people get to look under the hood, people get to judge it, make sure it's all there. There's going to be a dev net sometime soon, which sounds like a permissioned test net. So like let some early people in there, start to try and break things. Then an actual test net comes later after that. And so public test net, permissionless test net, everyone gets to come and try and break this thing or start to build their app or just do something. And then after that, kind of like the Ethereum test nets, after that gets like tried out in the wild for a little bit. We get more data, we get more feedback, we start to like polish off some edges. Then we'll go to mainnet and now we have a fully public permissionless ZK EVM on mainnet.

• 11:40Speaker 01

  Exactly, exactly. And just, I guess, relatively important thing to note there is that what is ahead of us now is, I would say, a standard software, web3 software release process. There are no engineering challenges, no cryptographic challenges. This was the challenging part, proving these complicated dog codes, cracking that whole thing up, the whole EVM, basically, the whole implementation. That was the tough part. Now is basically just a more or less routine process of reaching the mainnet itself.

• 12:16David Hoffman?

  Well, that's certainly exciting. We've been talking about on Bankless, I've already brought this up in a number of interviews, but we're in this bear market, but we're not going to get out of it until we build our way out of the bear market. So we must build and like a ZK EVM can start to like produce some of the UX challenge, solve some of the UX issues at the top level for like, you know, instant transaction speeds. Basically, like zero transaction fees? Is that fair? Like, what's going to be the?

• 12:45Speaker 01

  When you start discussing transaction fees, you always have, you can technically allow zero fee transactions, but then you have this spam problem, which was why the fees were introduced in Ethereum in the first place. But we are definitely discussing negligible transaction fees. You can easily, I guess, compensate the transaction fees for users, but then again, you have this spam issue that maybe can be also addressed. It's a thing to discuss and consider, but.

• 13:12David Hoffman?

  Certainly. So like, if there isn't compensated fees, like I'm the user and I'm going to pay the transaction fee and I need to send you some Ether, can we guesstimate how much that can cost?

• 13:23Speaker 01

  Currently, for us to actually produce, when you take everything into account, the computing power that you need to actually generate the proofs, when you take all the. Running, I guess, operational costs of this solution that we have. It's currently, I can give you the exact figure, but it's 0.0087 dollars.

• 13:44David Hoffman?

  0.00, so one...

• 13:46Speaker 01

  So 0.00, less than one cent, basically. Is less than a penny is the actual operational cost that we have to, you know, process the transaction, to generate proofs and everything. And again, like this was, it's also worth noting that this is the first iteration of ZKVM. We have clear idea and clear path how we're going to improve the efficiency of ZKVM. It was really just, it was about proving that this is possible. Basically, because no one was certain, even us, you know, in certain, at certain times, we were questioning, you know, is this really too hard of a nut to crack? But it was really important to prove that it is actually possible. Now, when we have proven it, there are at least four major ways how we can improve

• 14:30Speaker 01

  and how we are improving ZKVM. Improve the cryptography itself. We have very, very strong cryptographic team now. Improve the code itself, optimize it, use optimal programming languages, etc. Pipeline the provers. When you have a prover that is, that takes one minute to generate proof, you can pipeline 20 of them and then instead of one minute, you have three seconds. And the fourth thing is actually hardware acceleration. We're also working on that very actively. That's not publicly known, I guess, but we have both GPU-focused and FPGA-focused efforts to accelerate on the hardware level itself. So there's four major ways how we can further improve ZKVM. But current cost to actually process one transaction, improve it, is, as I said, a little bit less than one cent.

• 15:15David Hoffman?

  Well, yeah, so the base resource cost is less than a penny. So like, maybe there's like a, because the fee market gets to like a penny or so.

• 15:22Speaker 01

  Yes, it's up to market then. Like, it might reach two pennies, three, it's, it depends on the market.

• 15:28David Hoffman?

  Two or three pennies, oh no. It's what it can be.

• 15:30Speaker 01

  It depends on the market. I mean, yeah.

• 15:33David Hoffman?

  The other really exciting thing you said is the EVM equivalence aspect of this, which is something that I went down that rabbit hole and came out very, very appreciative of what EVM equivalence is. And just to speed run the listener through that, you have EVM equivalence versus EVM compatibility. And EVM equivalence is like the true copy and pasting of code without having to change anything. Like the EVM stack... Is one-to-one comparable where EVM compatibility is like, well, the bytecode works. I might be fledging this, but like the bytecode works, but like producing that bytecode is like a little different and like you need a different developer tooling. So with this EVM equivalent status of this EVM chain, you get to just port over

• 16:13David Hoffman?

  the developer ecosystem, like whatever that means, like compiling tools and like solidity tools. I'm a little bit like out of my water, out of my element here talking about this, but like I kind of get the gist.

• 16:25Speaker 01

  No, no.

• 16:25David Hoffman?

  And so like this is, this is coming out, like when you launch this ZK EVM mainnet, this is like you have everything else that all the other developers are familiar with on other EVM chains as well, like out of the box.

• 16:37Speaker 01

  Yes, you're leveraging the whole ecosystem, which is very, very important as we all know. So what you, what you're doing, we're discussing EVM equivalence. You're basically re-implementing the EVM, the way it should be implemented. It's like, almost like you're implementing a new Ethereum node, like, like yet, like you literally take the official EVM specification and you go op code by op code. You're literally re-implementing the EVM, but this CVM is ZK friendly, so to say. In the background, it is able to generate ZK proofs for batches of transactions. Otherwise, it's a normal EVM. It's implemented like GetEVM, like ArigoneEVM. It literally follows the same specification. That's one thing.

• 17:19Speaker 01

  You, you just implement all Ethereum RPC methods. Currently, you're actually running Ethereum, official Ethereum tests. There are thousands of tests that every client, for example, that joins, that wants to join the network should actually pass. Currently, ZK EVM is passing, the client's passing around 30% of the tests, but we just started, like we're very confident that we, that the percentage is going to go up very, very quickly. So basically, if you implement all the RPC methods, you follow the official EVM specification, you pass all the Ethereum tests, you're very much Ethereum equivalent. That's it. That's what you want to ideally have.

• 17:55David Hoffman?

  Right. And, and just to really illustrate why this is so powerful is that if something is EVM equivalent on one layer two, or on the Ethereum layer one. You can instantly just take that whole thing that you've built and put it on any other EVM chain instantly. So like all the apps that are built on any other EVM equivalent chain work on all the other EVM equivalent chains. So like we get, we get, there's tailwinds in the development labor when you work on an EVM equivalent chain.

• 18:18Speaker 01

  Exactly.

• 18:19David Hoffman?

  Yeah. So what's the name of this thing? I know like there's always like a name for every single Polygon chain. What's the name of this one?

• 18:25Speaker 01

  Uh, it's still being discussed. You know, like also one, that gives me a good, that's a good segue. Polygon Hermes officially shipped ZK, ZKVM.

• 18:36David Hoffman?

  Okay.

• 18:37Speaker 01

  But what is very, very important to understand, and this was really a collaborative process of almost one year of very exciting collaboration between Polygon Hermes, Polygon Zero, and Polygon Mid. And there was literally zero competition between these teams and 100% collaboration. And it was really amazing. We have regular, we still have regular weekly calls between the teams, daily asynchronous communication. Every month or two, we have in-person workshops, basically somewhere around the world. So it's, for me personally, it was really a, an amazing experience watching these, these smart people collaborate. And the reason I'm saying this is, that's also Polygon Hermes people feel very grateful to these other two teams for all the help and contributions.

• 19:21Speaker 01

  So that's why we're discussing currently the name that would maybe unify or symbolize that really this was a team effort. Yeah.

• 19:29David Hoffman?

  I remember ECC last year, Mihailo, it was you, me, and Mariano, and we were getting drinks after the conference was over. And then you went and like disappeared for like five or 10 minutes to take a phone call. And then you came back and then sat down at the table just super excited. And you wouldn't, and you couldn't tell what, me and Mariano, what it was, but then like a couple of days later.

• 19:48Speaker 01

  I told you, or I couldn't.

• 19:49David Hoffman?

  You, you, you said it was something about we acquired a ZK, or a, a, a ZK team. But, but you couldn't give us all the details. And then like three days later, it announced that Polygon had acquired Hermes. I was like, oh, that's what, that's what that call was.

• 20:03Speaker 01

  Yes, yeah, in retrospect, it's, yeah.

• 20:06David Hoffman?

  It's crazy how much has happened in a year.

• 20:09Speaker 01

  Yes, yes, very, very crazy timeline.

• 20:12David Hoffman?

  And so, so with all Layer 2s, there's always like a staking, a token staking component. Maybe not always, but like that's kind of where we assume this is going. And I'm assuming Matic is the token that gets staked to validate this chain.

• 20:23Speaker 01

  Yes, and currently also another, like, there's so many things happening in Polygon at every single moment. We are currently redesigning the token. That is something that is also actively happening as we speak. Basically, the Matic ticker is probably the last legacy of the previous project.

• 20:42David Hoffman?

  Right, because Matic doesn't make any sense anymore, yeah.

• 20:44Speaker 01

  Yes, and we intentionally kept it that way because when we announced Polygon, it wasn't just a mirror rebranding. It was really a symbol for us of this pivoting the vision and this new vision, grand vision that we had. And we wanted to do the same with the token. We intentionally kept it the same ticker and the same design. Until we were actually confident that Polygon as a project and the vision has product-market fit. And now we are ready to actually migrate and upgrade the token actually to something that will reflect where Polygon is. And then we have made certain adjustments that will make it, I guess, a powerful utility and a very important component of the Polygon's wider vision.

• 21:27David Hoffman?

  Amazing. Well, Mihailo, what are you going to announce next year at NextE CTC? Every single year, there seems to be something.

• 21:34Speaker 01

  That's too, too long of a framework. We can talk in two weeks. Yeah, it's genuinely like hard for me to perceive like what would be the next.

• 21:44David Hoffman?

  What to do in 12 months. Awesome. Well, Mihailo, thank you for coming on and sharing the story.

• 21:48Speaker 01

  Thank you so much, David, for following our work always.

• 21:51David Hoffman?

  Of course, of course.