Build with Nethermind - ETHGlobal San Francisco

what's up everybody I'm John Church a product manager at nethermind and uh I'm going to be talking about the meev infrastructure Bounty um so um we're supporting a uh uh Stark net me infrastructure Bounty um because Stark net needs me infrastructure and especially as Stark net moves towards a decentralized ecosystem um and and we think that helping people get involved with some of the different bounties and and building products and ideas in the meev infrastructure space will help um Stark Net's uh ecosystem be more healthy and have healthier markets so um real quick high level what is me it's uh it's called Uh maximal extractable value value originally it was minor extractable value um this term has changed over time but it roughly refers to um uh the value that a minor validator block producer can extract from from manipulating or or um managing or messing with the order inclusion exclusion of of blocks um and transactions within a block um me can be positive me can be uh uh negative uh and that's why this is an me infrastructure um uh Bounty which is we want to help daps and the large your ecosystem make me a neutral or positive force because when me is properly accounted for properly built it becomes Revenue stream for apps and it becomes a revenue stream for the larger ecosystem and ultimately helps um make markets more healthy helps make uh um pricing in the markets like a Kubo and others uh makes pricing more accurate and aligned with other marketplaces like binance coinbase ethereum Etc um so it's it's critical critical for for supporting healthy markets and that's what we really want to enable um uh due to Stark net currently Stark net is a first uh first in first out um uh or first come first serve um uh uh transaction order flow so due to that the um me is is a is a bit Limited in terms of um uh which is good this is fine um you know it's front running sandwich attacks Etc are are very very hard to do um uh and that's great and that's partially why we wanted to do this as a stark net infrastructure um rather than anything else so just kind of highlighting that we want this to be to support healthy markets um so uh examples of of of me infrastructure um solver markets like cow swap or Unis swap X like these are some ideas that you can use to think about um uh uh you know what what would you like to build so solver infrastructure examples on L1 or cow swap where they created a Marketplace for solvers and um uh um user intent so user intent comes in I want to swap eth to usdc or usdc to eth um and then solvers go find the best uh uh routing and best pricing and then they present that they take that Swap and then they present that to the user so solver markets um dap infrastructure to inte integrate meev um this is a um something that I think you know cow swap specifically created the UI and then the the solver infrastructure I think that there's opportunities in terms of creating solver infrastructure that is UI agnostic and so you can plug into different um uh daps or daps could use that maybe like an DK or libraries um I think there's there's a lot of opportunities there um also generating intent markets and this is this is kind of an adjunct to the previous example of the the DAP infrastructure to integrate meev um generating a marketplace where solvers can compete where user intents come in um and then ultimately I think uh uh exploring this space is a big opportunity like this is the me uh infrastructure space is why wide open there's a lot of research a lot of thought a lot of Novel exploration around different ideas happening um in the the me infrastructure space how you aligning meev to um uh uh the different layers that they're on to the daps that they're on to the wallets that are being used I think that um there's you know new and novel ideas are certainly welcome here and especially uh me in relationship to um cryptographic like ZK integration in TV I don't think this has been explored yet um I think that there's a lot of opportunity like proving compute through ZK um I think there's there's some novel thought and novel products and and ideas that can happen there um just high level requirements for for judgment um to cons be considered your project source code must be hosted on a repository like GitHub must be your own work and you must demonstrate that it's executable on Stark net test net so it's got to work um your project will also be judged on the quality of the the cryptography underlying the project [Music] um basically that that um like how how much do you use the novelty of Stark net and ZK and and what's available um also the quality of the implementation will be judged and the amount of work that you've done on this during the event um and these are some resources that I think are important um and I'm just going to talk through them really quickly um the Stark net me on uh this I school's um Stark net M repository is just is a list there's a number of uh bot Builders um bot templates and and highlevel um information on how does stark Network how is it different than uh L1 and um uh you know so if you need some quick and quick information on how to to get there that's a great place um Al the ethereum docks on me which is just a a highlevel you know primer on what is me how does it impact L1 um I included anoma anoma as a research resource because anoma's uh is doing some really interesting work in terms of um intense marketplaces which ultimately you know capturing intense and moving them into um uh solver networks or or Searcher networks I think or in Searcher marketplaces I think is there's there's a there's a wide open space for Unique Products um kyang is of course the the the language of of Stark net so you'll need to know that and then the CR course in Stark net is is a publicly available um uh repos uh notion page for um that we've produced to um help people get up to speed on Stark net [Music] um I think that's about a wrap uh again I'm John Church best luck out there and get those bounties test test am I on okay cool hey everyone uh thanks for coming this morning bright and early um I'm Josh I'm a software engineer and engineering manager at nethermind and today I'm really excited to talk about monomer collaboration between nethermind and the amazing polymer team um to bring the cosmos SDK to the op stack before we discuss what exactly that means and how it works um I'd like to just get get a show of hands who here has like a would consider themselves a smart contract developer like you've written solidity or some other smart contract language okay good number of you um who here has deployed a smart contract on an L1 like ethereum mainnet or testnet testnet count okay basically everyone who was a smart contract developer um who has deployed a contract on an L2 testet or main net okay about half the number okay and then who here has deployed an L1 Network or an app chain as an L2 L2 okay okay actually more than I expected um but it was a different group than the the smart contract developers so the the purpose of of this talk and the purpose of monomer is to really bridge that gap between the smart contract developers and the maybe infrastructure developers um give the smart contract developers the the power of amazing infrastructure and give the infrastructure devs the ability to write um amazing applications um here let me grab the clicker awesome okay yeah so we'll discuss like the motivation for monomer we've already kind of discussed that a little bit um then we'll go into how it actually works um what it does how you can use it and then finally conclude with uh what's next for us so uh today developers um myself included have to choose between kind of this two polar uh two two ends of the spectrum you either have to choose like security and network effects if you want to deploy your application on like an L1 on ethereum Manet you get great composability um it's really easy to use you know metamask works out of the box you don't have to worry about all this underlying infrastructure it's kind of like deploying in versel and web 2 you know everything's taken care of it's a very smooth experience um and then on the other end of the spectrum you have like the equivalent of deploying on Prem your own Hardware in web 2 which would be like deploying an L1 app chain you know it's very expensive you have to um you know pay your validators somehow that includes inflating your token um which can be bad for your ecosystem and caring a lot about things that don't have to do with the actual business logic Logic the actual product that you're building um but you get some benefits it's maximally customizable so you get great technical sovereignty on this end of the spectrum you got great security great Network effects um on this end of the spectrum I think what's really interesting and where we should uh converge is somewhere in the middle of this U maybe you can compare this to AWS and web 2 you know maybe I'm stretching the analogy too far right now I I think this means deploying something like an evm L2 you see some chains doing this um world chain recently announced that they're launching a separate chain um a has indicated that they're going that direction as well lens um is another example so so I think deploying a custom L2 gives you some benefits of this you you inherit the security of ethereum mainnet um as well as maintaining some technical sovereignty over how you allocate blog space um and how your application runs under the hood so the main thesis of this talk is that this is kind of a false dichotomy there's no reason why you can't have technical sovereignty and control over your application while still inheriting the security and network effects of the platform that you're building on in this case it'll be the op stack and ethereum mainnet but I think this will apply to a lot of other ecosystems as well monomer focuses on ethereum Main net but I think the principles that we'll discuss today apply to um other chains as well um okay yeah so the thesis is you don't need to choose between sovereignty and security um you can build Cosmos SDK applications on the op stack with monomer and this gives you really really nice benefits that we'll discuss um toward the end of the talk okay so I want to just discuss H how this is actually accomplished because it may not be immediately clear how we take something from a completely different ecosystem Cosmos and deploy it on like the quintessential ethereum product um op stack so who has seen a picture like this before this is just ethereum's highle architecture you know the division between the execution and consensus layer raise your hands who's seen something like this okay pretty much everyone um and even if you're not in The Ether ecosystem you've probably seen something like this before because it's quite common most uh blockchain infrastructures converging on a separation between execution and consensus and this even exists in like web 2 applications where you divide like the control plane and the data plane is another way to to describe it um if you haven't seen something like this before at a high level in order to run an ethereum validator um you need at least two pieces of of software you need your consensus layer client you need your execution layer client they communicate using the engine API um with each other and then with other nodes communicate over some P2P Network um to actually add blocks to the chain and the way this works is the consensus layer decides which blocks will come next in the chain and the execution layer executes those blocks okay and it's really nice having the separation between the consensus and execution layer because as we'll see monomer actually exploits this to kind of mix and match um and do creative things by replacing either the execution or consensus layer okay but the really important Point here is that this division exists and uh the way that this and kind of the glue that ties the execution and consensus layer together is this well uh standardized engine API um and the engine API isn't like the details of it aren't really um important for most of this talk the important thing is just to know that this standardized API exists okay who's seen a picture like this before everyone's hands should go up it was on the last slide um in the cosmos ecosystem you have a very similar division between the execution and consensus layers um instead of communicating over the engine API they communicate over the ABC the application blockchain interface um there's some slight differences between the engine API and ABCI that we'll discuss a little bit later but the important thing to note is that this is almost a carbon copy of what we saw on the previous slide um the so in blue we have the C instead of well here let me just show you so in e theum um there are multiple implementations of the execution layer multiple implementations of the consensus layer in Cosmos um there the way things have developed there's usually usually your application is developed using the goang cosmos SDK and the consensus layer is comma bft um it's just an implementation of tender in blue we have the Cosmos syk app and then below we see common bft but they perform similar functions to the execution and consensus layers in ethereum okay so now um this is a little bit more complicated but it's the op stack shares a similar design to both the ethereum and Cosmos um architectures you have the execution layer which in this case is op and then you have this kind of more complicated consensus layer that actually performs the rollup functionality um and there are other components of the op stack that I'm not showing here you can think of this as like the the essential gist of how the op stack Works um a little bit of History the optimism so the op stack was originally developed for optimism specifically um optimism is you know a roll up on ethereum and a few years ago one of the primary selling points of optimism now it's more common in part due to the optimism team's amazing work is uh ethereum equivalence opt the idea was you could deploy your smart contract on optimism and get the exact same user experience and functionality that you would get on main net with just lower fees um and the way they did this was by taking the execution layer client from ethereum making minimal modifications like very very few and hey guys my name is aush and I work as a devil at nethermind and I'm going to tell you about the nethermind Bounty and also tell you about stocket and why stocket is a game Cher in the blockchain space and why you should explore it being in web 3 to understand why stocket is exciting let us first examine etherium score scaling challenges so that would be like the block size time constraint so with a 12C block time and gas limit there's a cap on how many transactions that can be processed on chain then comes the state growth as more transactions modify ethereum's Global state it becomes increasingly resource intensive for nodes to manage then comes the consensus overhead in ethereum all full nodes must process every transaction to reach an agreement this Universal validation requirement fundamentally limits the scalability for L1 then comes the data availability costs storing all transaction data on chain is really expensive and this leads to Rapid chain growth these factors together combined to create a network congestion and high gas fees on L1 especially during Peak usage of the network this is where two solutions like stocket come into the picture they address these challenges in four key ways first is increased throw put layer TOS process more transactions per second by moving operations offchain second is reduced cost the lower fees of the transactions by batching multiple transactions into one L1 interaction third comes the maintained security layer tws inherit etherum security through cryptographic proof generation which in case of stocket is specifically called called a validity proof finally and perhaps the most exciting one is that they enable us to do complex computation on chain this means we can now run resource intensive operations cheaply opening up entirely new blockchain use cases which still needs to be explored so starnet as a ZK rollup takes these benefits even further ZK stands for zero knowledge a cryptographic method that allows one party to prove another that a statement is true without revealing any information beyond the validity of the statement itself in Star Net's case this means that we can prove that a batch of transactions is valid without processing each transaction individually on ethereum this achieves greater scalability while maintaining security enabling a new class of blockchain applications which were previously INF feasible due to cost or computational constraints on ethereum so stet brings ZK rollups to life using stocks which is scalable trans argument of knowledge here's how it works actually so stocket represents its entire State using a Merkel petricia tree processing transactions offchain to update this state it then generates a stark proof a cryptographic guarantee that these updates are valid this proof along with the compact state commitment is sent to ethereum for verification the beauty of this system is its efficiency a single proof can validate thousands of transactions which dramatically reduces the data footprint on ethereum this allows stocket to achieve throw put and lower cost while inheriting ethereum's robust security features by moving complex computations offchain but anchoring the results on ethereum starnet opens up a new possibility for blockchain applications it's not just about scaling existing use cases which are present on different layers already but it's about enabling entirely new ones that were previously impractical due to cost or performance constraints now let us dig a little bit more on how stocket is classified so starnet is classified as a type four ZK evm it means it's optimized for efficiency rather than exact ethereum equivalence it's like we are reimagining what's possible with blockchain computation this approach was crucial for scaling ethereum because it allows for greater optimization and efficiency by moving away from the strict evm equivalence stocket can leverage custom built system that are specifically designed for generating zero knowledge proofs the tradeoff is slightly different developer experience compared to how you do things on ethereum however this sacrifice in immediate familiarity enables us to push boundaries of what's possible on the blockchain we can perform more complex computation handle larger data sets and create more sophisticated smart contracts in Cairo which is the programming language for stocket smart contracts all while maintaining the security guarantees of ethereum and this is why type 4 is so exciting it's not about making ethereum faster or cheaper but about fundamentally expanding the capabilities of blockchain Technology let me quickly walk you through starets transaction life cycle it's a bit technical but understanding this process is key to appreciating star Net's efficiency so the first step is transaction submission a transaction is sent to a Gateway entering the mempool with a received status the meol is like a waiting room for transactions holding them until they are ready to being processed then comes the mempool validation step the mempo performs initial checks like ensuring sufficient balance for fees this is similar to ethereum's Signature checking but stocket adds custom validation functions for different transaction types then comes the step three which is sequencer validation the sequencer revalidates transactions before execution ensuring they are still valid in the current state the fourth step is execution valid transactions are applied to the state failed transactions are included as reverted The Next Step would be proof generation and verification this is where star Net's magic happened the Prov generates a cryptographic proof for the entire blocks execution which is then verified on ethereum L1 this process allows stet to batch thousands of transactions into a single proof significantly reducing costs and increasing throw put while maintaining ethereum security it is this efficient proof system that enables more complex computations possible on stocket one of stock Net's unique feature is its built-in account abstraction unlike ethereum's eoas which are externally owned accounts all accounts on stocket are smart contracts this enables signature abstraction which means that you can customize how your account validates transactions then potential for fee of substraction which means that we can pay gas fees in different erc20 tokens then comes non- substraction which helps us to send multiple transactions in parallel so this account abstraction built-in feature in stocket opens up possibilities for various applications like social recovery wallets or biometrically authenticated wallets for your dap so stocket uses a programming language called Cairo now the question comes why do we have to learn Kyo by a different language so let me introduce to the features of Cairo it's a language which is designed specifically for creating efficient ZK proofs which is the key to Star Net's performance Ciro allows developers to write complex logic that can be efficiently proven enabling complex computation that were previously in feasible on chain you can do all these things just by writing smart contracts in Cairo you don't have to create any more circuits or anything like that for for creating ZK proofs now if we talk about how you can dive in and start learning we have got a range of resources to suit different learning styles and experience level first there's an official stocket documentation it is your goto place for up toate guides and references for anything related to stocket it can help you start by setting up your environment then declaring your smart contracts creating an account deploying your smart contracts and then also interacting with them you can deep dive into the architecture section to understand how stocket works from an architectural point of view for a deep dive into Cairo itself you can use the Cairo book it's a comprehensive guide that takes you through the language step by step to create provable programs on stocket as smart contracts or even a standalone provable program is possible to be written in Cairo so this would be the goto place to learn Ciro from very basics for people who actually like the Hands-On approach for learning any language I would suggest using stock links where you can understand how smart contracts work on stocket by going through all these exercises exercises and compiling them and also seeing if things work the way you think they are working other than that we also have Cairo on exorcism which can help you learn Cairo for those who enjoy a more gamified experience for learning anything node Guardians would be the best place because they provide challenges and quests to level up your stocket skills we also have a crash course on stocket which is a list of various resources that you would use during your application building on starnet so I would walk you through all the resources that I mentioned here first would be the open RPC playground this will help you understand what are the different methods that you can call from a stet API for example like the stocket spec version which Returns the version of the stocket Json RPC specification which is being used now the question comes how do you create these code Snippets if that is also something which you want from us then you can go on to the crash course page and then migrate to the stocket Json RPC Builder this is where you can request for code snip PS for interaction to stocket using stocket JS or stocket go or stock net. RS for example I'll walk you through a request which we can create here to get started here I would need an API key that I can get from Voyager which is the blockchain Explorer for stocket now we can go to access stocket RPC get started for free and here you need to create your own RPC API key uh I'll name it as demo then we can copy this API key and go back to the Json RPC Builder this one and paste the API key here now I want to call this particular function starnet spec version and I can just click on send request you can see I get a response here saying that the result that the spec version right now is 0.7.1 you can copy code Snippets from here if you're using star. Js if you're using stock.

go you can migrate here you will get all the code Snippets which are required for creating applications on stocket in the crash course you can also see that we have an introduction to Juno which is basically a go implementation of a starnet full node client which is created by nethermind this would allow node operators to easily and reliably support the network you can create your own Juno instance on a system and you can use that for putting the RPC request that you are creating in this Jon RPC Builder apart from that uh these are the things which you can also use like the stocket CLI then stocket arent architecture documentation which is also present in the stocket documentation then the cyob book and this is the bridge that you can use for ethereum to stocket so basically basically this bridge helps you migrate your assets from ethereum L1 to stocket within 3 minutes but a migration from star net to ethereum L1 can take uh maximum up to 7 hours then here are some links to the major daps documentation which are present on stocket iikubo is uh the largest amm present on stocket right now and largest transaction router is Avenue so you can use these dap documentations to build something on top of them as well then here's the link for the stocket ecosystem you can go to this to explore all other community-driven resources that are present for stocket whether you are a beginner or an experienced Dev looking to expand your blockchain skills all these resources will help you get up to speed with stocket and Cairo we have a plenty of resources apart from these which are driven by Community as well but to get the link of all these resources you can scan this QR code now here comes the most awaited part the discussion about the stock net. anything bounty we are offering a price pool of $1,000 for the most Innovative project that can be built on stocket so star Net's architecture allows for some interesting capabilities that W feasible before onchain so we have some ideas of what you could build around so things things like di AI verifiable compute privacy cross chain applications or games but honestly these are just suggestions what we are really interested is in seeing what you come up with maybe it's something even we haven't thought of yet that's the coolest thing about new tech right Sometimes the best ideas come from fresh perspectives or hackathons like E online the qualification requirement for stock. anything Bounty are pretty simple the project source code must be hosted on a code repository like GitHub the code must be your own work and you must demonstrate that your project is executable on star net test net and is working fine completely there we are not expecting you to revolutionize the blockchain overnight we are just curious to see what kind of Creative Solutions you can build using star Net's features so give it some thought we are excited to see what you'll create remember the goal isn't to make existing applications cheaper or Mig in applications from different blockchains to stocket it's to imagine and build things that weren't possible before so if you have any idea that you think could work very well on stocket we would love to see it and don't worry about being perfect or world changing we are here to help you for whatever need you have or whatever doubts you get so good luck and let's push the boundaries with stocket hello thanks for being interested in this bounty the best zkl project Bounty um so in this for for this project I we challenge you to build a a project that combin ZK technology technology and machine learning so what does this mean this is a very open-ended task you might consider oh I'm going to start again I'm a bit nervous I don't know why I feel like I'm talking to a wall okay uh thanks for being interested in the zkl project uh so the main task of this project is very broad it it it asks you to build a project that combines ZK technology and machine learning this is a very broad description it is very open-ended so um the goal of this short presentation is to give you some ideas of how you can tackle this this project so you might want to start focusing on a sub branch of zkl so there are several sub branches there's one of the sub branches is proving correct in inference of a machine learning model this means that you prove that you execute a machine learning model that's maybe signed by a trusted Authority and then you create a proof that this model produce such an outcome on a public date on a public input and you also prove that the model that produced this output is signed has a has a correct signature but by this trusted Authority this is a very hot application of zkl it basically allows someone who run the a model to provide a very suing proof that the model was executed correctly and also you can prove that your model is signed by some trusted Authority so for example maybe the model is signed by the government and suddenly you have a proof that is succeed very to verify and this proof says that you pass some some criteria and and uh you you are you're also proving that the model that the government signed this model so you are providing a proof that the government thinks that you pass this this criteria and yeah and there's there's several aspects in here there's the sness so the proof is a very short it it can be very easy to verify as compared to just running the model you may also want to explore the ZK part of ZK ml for this application maybe you want to hide the input as well or or parts of the model who knows um surprise us there's a another subran of zkl which is a bit it is a bit less explored less developed because it is significantly more complicated to in in practice to achieve than correct inference and it is proving the correct training of a model so there's some data set and you want to prove that a model was train on the data set that's quite complicated because the training of a model is as you know the most expensive part of ml it um so training a large language model can cost billions of dollars or Millions um yeah so proving such a computation with on asnar is is looks very out of our Rich right now but it is certainly an area that is being researched and and there are certainly teams that are tackling this problem so you might want to also take a look at this there's another branch of zkl called Federated learning in Federated learning there are several uh parties that have data and they don't want to share the their data because it's private maybe let's think of let's think um a network of hospitals each hospital has its own data they don't they don't want to share the data with the other hospitals because uh they just cannot legally so but but the hospitals want to train or run some machine learning model on on all this data and they want to merge all the data together so they they have a a more um a more uh deep data set so Federated learning aims to solve this problem because they cannot just share the data databases and merge them together because of legal reasons they would be uh disclosing patient information to to third parties uh so Federated learning is a branch of of machine learning or dkml that um focuses on allowing this set of parties that have their own private data set to train or run models on this on this set without revealing the data sets to each other and yeah so these are some suggestions we strongly encourage you to to think of your own projects and and to surprise us with with your cre creativity in fact creativity is one of the judging criteria of this of this bounty we we want to see so of course we want to see projects that deal with zero knowledge machine learning but we will value in high we will have in high regard projects that are very creative that that are built upon an idea that it is surprising and and interesting so yeah and to finish the this uh short introduction to the Bounty let me give you some suggestions and recommendation so it might be easy to start at by looking at the problem of proving zkl inference that rather than for example proving training because as I said proving training is a much much harder problem in terms of efficiency than zkl inference and also there's much less research done there however as I said we B we value creativity highly so if you are interested in um in exploring exploring the training problem we will we will also be happy with that and we will take into account that the training problem is is harder than the zkl inference we also recommend you to try to look and use existing Frameworks for zkl or uh existing uh domain specific languages you one way to start a project might be to look for some existing application or project and try to optimize some part of it for example one part I believe needs some some attention in zkl is how the floating Point arithmetic is treated right now so I think that many projects uh prove operations with floating Point arithmetics in ways that are a bit more expensive than they could be I think this could these operations could be improved with uh some Lookout operations for example some yeah basically Lo up Lo up operations maybe lass of source decomposability Etc so you might want to go to some existing project in ckl and identify some some some part of it that's non optimal and and try to come up with a a small Improvement there that would be a great a great great way to tackle this this bounty and also please feel free to approach us and ask us what what are these uh pain points in ckl that you could focus on we will give you more examples like the one of the like the floating Point arithmetic exam example yeah so also don't be super ambitious like always remember that the hakon last what it last a couple days this this can be a hard problem getting into a very big project might be might sound great but there might not be time to to do that in a couple days and yeah please look at um at some possible useful links that we added in the in the description of the bounty good luck and let us know any questions you may have we will be around