Welcome to East to E to E to E to E. Okay. Hello everyone. Before we start the talk, my name is Victor. This is Leo.
In case you haven't noticed, we're actually brothers and we also work at Starkware. And the goal of today's talk is to explain to you how Starknet is going to scale both E. It is scaling Ethereum, but also it's going to scale at the same time Bitcoin. And so, um, to give you a representation of how we're going to do this, we have some hats so that you don't forget who is scaling what. Oh, thank you.
Testing. Okay. You're going to be Bitcoin today. So, I have a Bitcoin hat and Victor is going to be wearing an ETH hat. All right.
And then we also have like two hats here, which is Starknet's nextG prover, but this is for later. So, I'm going to hide this over here for now. But yeah, the goal of today's talk is to explain to you why we want to scale both Ethereum um and Bitcoin and um a high high level overview of the design choices for dual settlement on both Ethereum and and Bitcoin. So why do we want to scale Bitcoin? Bitcoin as of now is is the the most used blockchain out there.
It's um next to Ethereum. It has the highest level of of assets on chain, but it's not really usable at the moment. Um, that's just because it's it's purposefully built such that it's highly secure. But in a way, this turns the the chain into um a network that is difficult to use, that is uh often times very expensive, often times very slow. And the goal with us, Starknet, a layer 2 settling both on Ethereum and on Bitcoin, scaling both networks through one network is to give access to um to give access to DeFi to 1 billion users.
So let's go straight into the questions because there's a lot of tribalism between Ethereum and Bitcoin and other chains. So, Starnate is trying to unify these two chains or um settle on both. And we're gonna answer a bunch of questions today to why is this good for Bitcoiners and Ethereums? And I who is like a hardcore Bitcoiner here. We have Talip from Citra.
Okay. I guess everyone else is here for Ethereum. Who's a Who's a Ethereum Maxi? Okay. All right.
We have Okay. We have one Bitcoin maxi and two ETH Maxis. Guys, careful that things might heat up. But yeah, the this is Satoshi's white paper. The goal is to to have a peer-to-peer electronic cash system, and we don't have this today.
This is not possible to to have this um a UX friendly and cheap way of of doing peer-to-peer transactions on Bitcoin today. It exists, but it's difficult to scale. Yes. And you have solutions like Cashew, but this is Cashew is very centralized, which allows you to do Bitcoin payments. You also have the lightning network which allows you to pay with Bitcoin.
But these are alternatives and we want to have payments directly on the on the Bitcoin L1. So this is good because of scale. Now Starknet is a layer 2 scaling solution. It scales Ethereum. If you look at Bitcoin's max DPS, it's always 8.
5. And you cannot really bring um an international like um like uh a a chain that is supposed to to uh support um network state attacks um at 8.5 TPS. You cannot you cannot do that. And Starnet is a network on top of Ethereum.
And we've noticed a at the latest highest TPS 992 transactions per second. If you're just talking about a payment system alone, um, yeah, it's difficult to do on 8.5 TPS. Yes, he I say things and my brother does the TLDDR. Okay, this is also good in terms of cost.
Bitcoin is very first, it's very expensive and it's also very varying. So, it's it's not predictable if we're going to go from 1.4 to again 20.6 next week. But, Starnet is very predictable and it's also very very cheap.
you see transactions on on on starknet that settles at the moment on Ethereum at 0.001 um um dollars. So, it's extremely cheap and it's also extremely predictable, which is what you need if you want to create a D5 project or if you want to bring on institutions that need to know how much they're going to spend on transactions when they're using web 3 or when they're using D5. You can take the same principles that we have um for roll-ups on Ethereum um and basically unlock massive computation at a very very low and predictable cost that settles natively on the Bitcoin network. This is also good for UX.
I was saying earlier that doing transactions directly on the one is totally possible. I'm sure we've all done it, but it's really slow and it's very expensive. Often times it's unpredictable. So it's good for UX. Starknet has great UX.
It's this layer 2 that basically does offchain computation but takes care of the UX um for users. This is an example of a really successful D5 project on Starknet which is called Avenue. It's a DEX aggregator. And Avenue has built-in pay master so sponsored transactions. Users don't really have to pay for their transactions often times don't know that they're paying gas fees.
And it's also another cool feature is the the um the DCA order that it has integrated and you can just DCA into it and you won't it's it's the the UX is super smooth. It's not clunky at all. And so Starknet is this layer that takes the computation offchain. It takes care of doing transactions offchain and all it does is post a proof to Ethereum that whatever happened offchain is zk correct is mathematically correct. And on a bit of a higher level, once you have a layer that is able to natively settle to the underlying Bitcoin network um in a in a trustless or trust minimized way, then you everything that so you you unlock a level of programmability that you didn't have before.
And with that also comes things like you know um social social login, wallet recovery, all the sort of things that would let you engage with the network in a more friendly way that you that was harder to do before. Yeah, this is a feature that we have natively on Starket which is account abstraction. You can program your user's UX. If you want your user to see a 0x wallet during their um UX, you can just program that out. It's possible today.
We have an app called Focusry, which is a consumer app. It allows you to to you you you said, "Hey, I'm going to focus for 30 minutes and then you get an NFT for the time that you focus. Really super basic application, but it's on boarding 200,000 users in the last two weeks on Starnet." And there's the wallet is completely abstracted and the fees are um and the transactions are also sponsored thanks to Avenue. So that also means you could have a ton of users in the future using apps that are secured by Bitcoin without them even knowing that they're using the underlying Bitcoin network.
Yeah. And that's the ultimate goal. We don't want to give out gibberish to people about, hey, this is web 3 stuff. Like no, there's no web two or web 3. This is just consumer applications but with a web 3 background or back end.
This is also good for utility. So if you can have access via starknet to bitcoin's liquidity, you can stake your bitcoin, you can trade it, you can lend and borrow it. Now for the ethereum axis over there, why is this good for you guys? because now you have access to a much broader pool of liquidity and you can actually tap yourself into the Bitcoin liquidity and have DeFi applications that tap into this liquidity. But you also keep the custody of your assets that are on Ethereum.
So by being an L2 on both Ethereum and Bitcoin, you don't have to switch networks. Everything happens on one L2 and you can start tabbing into the the Bitcoin liquidity. And I think an interesting point as well is um like there are there are bridges right between between both L1's that exist but I think this could be the first or we hope this could be the first um version of the two networks properly talking with each other um and in a way sort of sharing and verifying state highly secure as well. That's the slide. That's the So why Starknet?
You know, why us? This is the track record that we have at Starkware. So Starkware is a company. It's a it's a it's a lab. We build infrastructure for Ethereum.
We've built a proof system called the Stark proof. It's been invented by the CEO, Eli Ben Sasson. And we have a pretty strong track record. Um, first of all, we say we are going to do things. We said we're going to like scale Ethereum with ZK.
And everyone told us we were going to we were crazy. Oh, you cannot write circuits. It's it's way too expensive. It's way too slow. Ah, everything's going to break.
But no, like we're the cheapest ZK rollup on Ethereum today. So, this is our track record. We invented the Stark proof system. We brought it to production grade. We built the first to the first L2 rollup, which is StarkX, which is which is more of a SAS.
So it wasn't permission. It isn't permissionless. Starknet is permissionless. Anyone in this room can build on Starknet. We have proof aggregation.
We then we we Starknet went mainet in end of 2021. We came out with recursive proving with circle starks which is this other way of doing cryptography with Starks which is really efficient which is also what allows us to um verify stock proofs on Bitcoin. We came up with we were the first L2 with L2 staking and we will be the um the the Bitcoin L2 that has stark verification on the um on the L1. So Starkware has a history of saying things that sound weird and strange and crazy and then actually doing them and then they become mainstream in the way that other people also end up doing it and totally normal. So I think the the takeaway is if we say that you know um there is a way to to to scale Bitcoin natively and to eventually have um you know a rollup that scales multiple underlying L1s at the same time.
It sounds weird and strange today and there are many problems and questions but there's a high chance that it ends up being uh ends up being the future. Yeah. Yeah, that's what we're building towards. There are big brains at Starkware and they're working on this right now. So the big question is how do you do dual settlement?
How do how does one layer starket settle on both Ethereum and Bitcoin? And now we're going to enter the end of this presentation which is a high level overview of how we plan to do this the design choices that we have thought about and that we also will publish very soon um on how we will um settle on both but this is just a high level overview and I think it's also the first time we do this publicly where we tell people the design choices but bear in mind this is high level overview we are not cryptographers or core engineers working on stark proofs and zk so if you want to learn more about artists do get in touch with us and we will point you in touch with the engineers working on this. So the goal is to bring 1 billion users onchain. That's the goal. What this is wait we missed the slide on dual settlement.
Okay. Yeah, this is so how does it work? This is starknet. Starknet produces blocks. Blocks represent starknet state and then we have bitcoin and ethereum as L1 as settlement layers.
There's a before we get into this, there's a couple of ingredients that you need for this to work. Um, and these are so you need a way to um basically so couple things. First, you need a light client on Starknet that for Bitcoin that is able to um basically tell Starknet what the state of Bitcoin is. You need the same on the Ethereum side. Um and then you need a way for on Bitcoin to verify um ZK proofs in a trust minimize or at least trustless way and there are various ways of of getting there and Talib from Citri can tell you more about it.
Um but yeah those are sort of the the core ingredients if you have those then what what we're about to say can work. Yes. And uh Bitcoin light client, we have this today. It's a light client written in Cairo. Cyro is a native language to build things on starknet.
Cairo is a language. Anything you write in Cairo gets compressed and can be very easily um verified um and very cheaply as well. Anything you write in Cairo is stark proven. In comparison to noir, we had Pedro earlier. Noir is also rust based, but it's for privacy.
So anything you write in Noir will be private. And so I had like this analogy the other day on Twitter. Um noir is the rust of privacy and Cairo is the rust of scalability. And you can use noir on starken but this is another conversation. Okay.
Um so we just produced the state on Bitcoin. Now we're going to uh post the state on both the the L1's. Sorry. We produce a state on Starknet. Yeah.
Yeah. Starket. sorry, created a proof of that and then sent the proof to both of the L1's right and so now the the state of startnet on both L1's is pending I think stuff is changing but oh yeah we move to state two but state one is still pending on both L1's afterwards um our Ethereum light client is going to notify uh starknet that the stark proof has been verified on the L1. And now if we want to have dual settlement, we need also the stark proof on bitcoin to be verified. So we consider these states as pending.
And once both uh L1's agree on the state, this is considered um final squared. So now the light client told starknet, hey the stark proof was verified on the on ethereum. And then the bitcoin light client is going to bitcoin is also going to verify the stock proof and it's going to say hey uh starknet the um the stark proof was verified on bitcoin. And then once there is verification on both L1's we have a final state for both. Okay.
What about reorgs? This is the last slide as well. So um reorgs what about reorgs? So reorgs are basically the um they're the elephant in the room which is okay if there is a reorg then what do you do? What happens?
Well there are a couple of edge cases and we're happy to dive deeper into them. Um it like let like come up to us later and and talk to us and we're happy to discuss them. Um but until they are there a couple of edge cases some of them um are fine and they can be fixed and others of them are more difficult to deal with. But the important thing to note if we go back to this previous slide is that in the happy flow both of these states agree and um and this is all good. And so the the question is, okay, what happens if either Bitcoin or Ethereum reorgs?
Um, and like I said, there are a couple ways to to to fix this, but the important thing I think to take over to to take home is that um these states will um so they can match, right? But they will also always exist independently. So the there will always be a state of starknet on Bitcoin and a state of starket on Ethereum. And so if something happens where these two states no longer matched and they cannot actually be reconciled, they will always continue to live on each of the L1's independently. So funds will not be lost in that sense.
you will always be able to withdraw back to the individual L1's. There are a couple edge cases again and we're happy to talk about them. Um, but what's important to know is that think of these as two planes and you're building a bridge between two planes that are also, by the way, flying at different speeds. Um but if that bridge breaks for whatever reason because of a reorg and an edge case for example and they cannot be reconciled then both planes continue to keep flying and um like startnet continues to to work and continues to post proofs to both of these L1s but you have a you would have a sort of separation of of state on starknet itself. Anyway, couple of edge cases like I mentioned, happy to talk about them in in in Q&A.
Um, but that would be that would be addressing the the reorg the reorg edge case. What's also important to know is that the fact that these two states exist independently is important if you are a maxi. So if you're an Ethereum maxi or a Bitcoin maxi and you're like I don't actually want my Bitcoin to touch any Ethereum state then that is that needs to be possible and so that's why these exist independently and you connect them. Yes, the design is ready. So if you want to talk to us and learn more about it um we can do it after this presentation.
So this is the end of the presentation for today. We hope you enjoyed the what we presented for you today. Um, and yeah, we'll catch you anywhere in the venue. We're here al also tomorrow if you have any questions. Um, also Leo does does BD and I do ecosystem.
So he works with big teams. I work with smaller teams and my goal is to help you find a Cairo job and or help you build your first project on Starknet. Thank you.
Automatic transcript — names and jargon may be misspelled.