Speedrun Rollups: A Beginner's Guide to L2s, ZK, and WTF People are Talking About on... | Devcon SEA

[Music] [Applause] okay hello hello thank you everyone for joining this is such a big room so I'm very nervous and I'm glad you're part of this journey um hi I'm Emily I lead the devell team here at scroll um and what I'm here to talk about is basically go through a laundry list of everything I've been nerd sniped by recently um are we going to head into the slides okay cool yes um cool so the title of this speedrun rollup ZK and everything the F people are talking about on panels um I guess brief disclaimer I'm going to talk about a whole bunch of stuff at a very high level um and if you have any questions I will try my best but this is more going over breadth than depth um to hopefully prepare you for all of the gigabrain stuff that's coming on later for the actual experts in the room so we're going to start from the basics L2 fundamentals dive into a little bit of ZK and then just get into sort of like current events things you might have heard about um to get ready for the rest of the week so starting from the very beginning like very very very very very beginning um kind of where did the Inception of rollups come from so if you guys have not heard of this phrase before the block blockchain trilemma is essentially a pick two solution between security scalability and decentralization uh So eventually effectively what that means is you can't have all three you have to pick two what that also means is the ethereum community has picked security and decentralization so what that looks like for you is what it looks like for this guy right um it's a big ux problem transactions throughput takes a long time that is a barrier to entry for kind of new people into the space so what is our solution you might have heard of this it's called an L2 so to understand what an L2 is let's take a step back let's start with the L1 so the L1 is kind of the B blockchain so if you guys know ethereum you know Bitcoin um that's an L1 so that that encompasses everything right that produces the block producers the node operators blockchain itself history etc etc layer twos are basically a separate blockchain that extends ethereum by pulling off the execution layer from ethereum um and what is special about it is it posts all the data back to ethereum and guaranteeing something called Data availability so data availability should set off a light bulb in your head this is a very important concept that will kind of thread through the rest of you know when you talk about l2s in the space in general um side notes so like little things just to keep aware of some stuff like other scaling Solutions you might have heard of in kind of the beginning was something like polygon proof of stake that is a side chain so side chains are also separate blockchains but it's a two-way Bridge so all of that information is still stored on polygon it is not brought back to ethereum so you don't have that same security guarantee um which which is why you know ethereum kind of started taking this L2 Centric road map um so let's dive in a layer deeper so l2s there's a lot of different solutions rollups have been kind of the canonical most supported one um how does that work exactly so like I said rollups take the execution burden off of the L1 what this looks like is hey I'm going to submit all these transactions to um I I work at scroll so let's say scroll right and what we do is we will bundle up all those transactions together and submit it once to the L1 so what you can think of this is is as like amortising the cost of all of those transactions um on the L2 side making it cheaper for everybody to interact versus doing all of them individually to the L1 um and so like I said again data availability is something we keep thinking about right because all that data is put back to the L1 reverting a rollup transaction means you have to revert the L1 itself right so that's how you inherit the security guarantees I'm going really fast so anyone like just like raise your hand if you're like Emily you're talking too fast um so I'm going to move on um so basically how do rollups work right so there's a few components to look at you have Smart contracts so you have contracts on the L1 that basically store rollup blocks um you can think of them as controlling the logic of the rollup for ZK rollup specifically there is a separate verifier contract to pull in and verify the the proofs I'll talk about that later um and then a bridge contract simply for message passing between the L1 and L2 the second piece of a rollup you might want to think about is the offchain VM um at a high level it just handles the execution computation of transactions uh away from the evm away from ethereum and the last piece is kind of an umbrella piece it's hard to kind of Define because every rollup does it a bit differently but you can think of the operator as like the guy who does everything um they're essentially things you might have heard of like the quincer the validator theover ETC um they are responsible for essentially combining those transactions like I mentioned earlier compressing them um and then submitting that block to ethereum so that's rollups at a high high high high level let's actually disect the rollup landscape right so now we understand the the what and the why of rollups now you've probably heard of a bunch of different types and how do we even decide what what uh what to build on what makes them different and so I think probably the first thing you've probably heard of is optimistic versus ZK Roll-Ups right so I think the thing to note specifically again we talk about data availability when we pose that back data back to ethereum how do we prove it's true right so there are two approaches optimistic rollups came first what they do is optimistically assume that data is valid so they Implement something called fraud proofs what this means actually is when we do something called withdrawals so withdrawals being bringing like eth from the the L2 back to the L1 they have to implement something called a challenge period so a challenge period is 7 Days basically where it's locked up and gives the community to opportunity to essentially dispute the transaction and so what that looks like is to say hey I want to dispute that within the seven days and then we will submit a fraud proof validate that fraud proof and then if it's good then we're good to go if we're bad then oh no you know um the downsides of this right is a few things right it requires at at least one honest node so obviously there has to be a good person I guess going through and doing the fraud proofs the other thing is this needs to post all the transaction data back to the L1 right because what it does needs to do is like recreate the history of everything going on um so that is something I will touch upon when we talk about ZK rollups right so ZK Roll-Ups was kind of the next phase of rollup architecture um this is when we think about okay rather than optimistically assume let's actually submit something called a validity proof upon submission of that data back to ethereum and so what this does is allows us to like create mathematical guarantees around validity um creating safer trust assumption security guarantees as well as when I talked about finality before with the challenge period you don't have to do that right you can submit it and then it's verified immediately um I think there are a few things I want to note here that I get a lot of misconceptions about ZK which will dive into a little bit later right is very popularly known as something we use in privacy with regards to rollups specifically depending on the ZK rollup privacy is not necessarily guaranteed so the use case of ZK proofs in this situation is posting that proof back to ethereum is expensive right every time we interact with ethereum that is expensive so we want to bring the smallest type of proof back and it just so happens because ZK proofs have the property of not having all the information revealed as part of that is kind of the most uh what do you call it uh smallest proof that's not correct English grammar but you get me right um to be put back so the uh last thing about that too right is like we only need to post like transition state data back to the L1 rather than everything as compared to optimistic Roll-Ups um so some examples you might know I put up logos scroll ZK sync you might know um what's it called Linea polygon zkm um lots of interesting players in the space uh but let's go a step deeper okay we've covered the optimistic Roll-Ups we've covered the ZK rollups but let's actually differentiate within the ZK rollup itself right how do we start thinking in that design space uh so the first thing I need to talk about is the ZK evm so essentially what you can think of a ZK evm as is like a magical computer that allows us to execute evm code so you can write solidity smart contracts but ALS also allows us to kind of translate that into something that is provable right because ZK proofs are math and how do you turn code into math I'm not entirely sure ask our researchers and Engineers but that's um what a zkm essentially allows you to do this is incredibly difficult to do just based on how the evm was built which has given rise to four different types of ZK EVMS so this is kind of where we started talking about okay how did the ZK Roll-Ups differentiate themselves from each other so at a high level type four is going to be something that we call High language level equivalent but not bite code equivalent so an example would be like a stark net or a ZK sync what this means for developers specifically is you can write in solidity but your tools will have to change to an extent right because it's not natively the same so if you're using something like a a hard hat or a Foundry you might need to install an additional plug-in in order to kind of understand the information type three I'm going to gloss over as sort of the transitionary period to type two which is kind of the next St AG I put a frog there because I like this one um but anyways this is a highle language equivalent so the same thing you can write in solidity you can write in Viper Etc but it is evm equivalent so you have bite code equivalents so what that looks like for your tools right is I just natively understand it there's not an additional overhead um you know I think this is a really strong factor that I like specifically is because you get the network effects of ethereum right something I always point out I think people don't think about often is let me stop here how many of you guys are familiar with a Viper hack Viper okay one one hand a few hands um let me take a step back uh this is a design space I think about too when it comes to ZK ebms so to talk about Viper if you guys aren't familiar Viper is like the python version of writing solidity smart contract or writing smart contract so not solidity but in Python um something that was really interesting that happened was there was a bug in Viper that caused a hack right and the problem was not in the Viper code itself but actually in the compiler right so if you map that same Paradigm to ZK EVMS right you can kind of audit all the smart contract code but there's a lot stronger at least in my opinion of a security guarantee when at the lower level you don't have that overhead of thinking how it's different as well so that's type two type one is kind of fully ethereum equivalent so again like I said uh highle language equivalent ethereum equivalent meaning um you can write in solidity tools don't change um and there's some little tweaks about like data and hashing that makes it the same but anyways the point of like also why you would choose different ZK EVMS is basically like I said in the beginning EVMS or ZK EVMS are really hard to build right so something like a stark net and ZK sync took the I guess stance of like you know evm is not perfect let's actually change it a little bit um in order to get certain vantages like maybe it's faster maybe we have native account abstraction built in versus maybe protocols more like scroll or Linea where we talk about no we want to maintain faithfulness to the evm because that's where the community is it just Smooths the developer experience things like that um everyone's super cracked and super smart so I'm not going to say one is better than the other but I think something to think about as devs when you're comparing the landscape um other things to think about because we love numbers when we talk about rollups um is the different stages so stages was something that was by L2 beat um if you guys aren't familiar l2b is a really great resource to look up like I guess rollup information that tells you about like different security guarantees um like their activity and stuff like that um they Define three different stages of rollup so again when you're thinking about the layer 2 landscape there's different um I guess what we call training wheels that we have as the technology progresses over time right so stage zero is where I would say most of the rollups operate right now um the most important thing to think about there is pretty much like the operators are centralized um so operators being like sequencers you might have you might hear that phrase a lot centralized sequencer right um the source code is available for you to reconstruct State stage one is kind of the next step right how do we let go of those trading Wheels a little bit more um so then that's when we bring in smart contract governance this is when we introduce something called a security Council right so Security Council is like a group of people who are going to basically arbitrate whether or not an upgrade might be made things like that there are specific requirements about like some of the security Council people should be on the team itself and some should be external as well right and stage two is like woohoo no we're completely training wheels off it's fully managed my smart contracts proof system is permission list Security Council is only there to basically um handle like upgrades uh sorry not upgrades soundness areas so if there's a bug like creating an uh Avenue to basically say okay are we going to upgrade this making sure we have a a a good amount of time for users to exit if they don't want to be part of that upgrade but also another kind of design space to think about when you're looking at the L2 landscape um and so to kind of summarize everything that we've talked about there's a lot of trade-offs that fit everywhere here um at the at the highest level I wrote as by highest level I mean it's the first bullet point uh finality is what I was talking about so when you think about challenge periods um in optimistic versus ZK rollups that might be something think about when it comes to liquidity fragmentation stuff like that um I'm going to anticipate a question uh in case somebody might ask like oh wait when I when I withdraw from like an optimism to ethereum it feels instant I don't understand where's that challenge period coming in um so just to address that point specifically you might think about bridges like third party Bridges basically having separate liquidity pools to allow you to give that feeling um as we all know bridges are the most like hacked space in the kind of ecosystem right now but you know I mean there's a lot of strong guarantees that are happening that I'm not a bridge by Builder so ask somebody else about it but anyways last piece uh second piece right like I said again security guarantees and Trust assumptions so whether or not you're optimistically like waiting for a fraud proof versus saying I'm going to submit a proof on um submission yeah validity proofs uh and to that extent thinking about throughput cost again there's a lot of different implementations that have trade-offs maybe if you're building a game you might want something that is uh faster and cheaper with fewer security guarantees and a different type of app um developer experience so this is a trade-off when I was talking specifically about the different types of ZK EVMS and what you're willing to kind of um I guess appetize not what what you have appetite for um decentralization open soft Source software community so this is kind of when you get into the ecosite s ecosystem side of things so how much do you care about those parameters um and like I said again Network effects so that not only affects like how connected are you to ethereum for example but in your ecosystem itself if you are a builder um what other builders can you interact with like kind of additive things so that is rollups at a very high level I'm going to scare you guys a bit rollups are not the endgame they can still be very expensive um why so like I said um oh I heard some Chuckles let's look at this again uh but anyways um let let's let's dive into maybe something that is a bit more of a current event right so like I said when I talked about data availability that's putting data back on ethereum but what do we remember about ethereum every time you interact with it it is very expensive so how do we get around that right so this is something called EIP 4844 which was introduced I think sometime this year um which essentially created something called blobs right so blobs are a separate type of data structure than what was traditionally known as call data so before you would post information to something called call data and call data is permanent and there forever right the thing about rollup data right is you're really only posting it there for like verification so it doesn't need to be there forever right so blobs are actually an ephemeral type of data storage it will like go away I think after 18 days um I'm not the most current on that act number but essentially what that does is allow us to use a different type of space the other thing that is really cool about this as well is it is a separate fee market so when it comes to competing for like I guess data on ethereum blobs are completely separate space so what that looks like or means for rollups is you are not as exposed to the fluctuations in the market on ethereum versus in call data versus blob space so um I guess the result is this is actually still not the most perfect solution that's when you might start hearing things about like sharding dang sharding things in the future of the theorum protocol but just think of this as again a highle primer on what you can expect for the rest of Devcon um so other things that I just want to kind of say uh to put words in your head um is there are uh what do you call other scaling solutions that have kind of produced or been created so you might hear something called a validium so so a validium is basically a chain that does like offchain computation proves via validity proofs but they use offchain da so another kind of buzz word to think about is now we're talking about data availability layers I don't work for these companies so I can't go too deep into them but if you want to investigate Igan da Celestia ail are kind of the new design space you might think of there um an optimium I forgot an i is basically optimistic rollup plus offchain DM da and then the last one is going to be called a volition so a volition is essentially the marriage of a validium and a ZK rollup where you can kind of switch between whether or not you want to do onchain da versus offchain um if you're looking for specific examples again head to L2 beat.com it'll actually list a bunch of different projects in the space and also what they are okay cool um I don't know how much time I have left I figured I would be talking for a long time so ah ah okay my backpack is back there I have swag for somebody who shouts out out the answer here what the okay which type of zkm is bite code compatible ABCD okay those are all numbers and not letters I don't know the answer is H everyone is right or not everyone is right I guess but um okay this didn't work out as well as I thought I not I did not think there would be this many people here but thank you for for participating okay let's move on let's talk about zero knowledge okay so this is some knowledge about zero knowledge um basically at a high level right zero knowledge proofs allow for a PR so a person to prove to another person the verifier that they know something without revealing the information itself right so what does that mean specifically so when it comes to ZK proofs there are three main properties we think about completeness of the proof so that means if the prover knows the information they can convince the verifier with their proof soundness means if the prover doesn't know the information they can't trick the verifier and then zero knowledge is like I guess pretty self-explanatory the prover doesn't have to reveal any of the actual information itself so there are two types of ZK proofs that we're going to get into something called interactive proofs so this is a type of proof that requires back and forth between the verifier um a an example that is pretty common you can just Google eli5 interactive proofs and this is a picture you get from from chain link essentially what it is is this example says okay I have um Bob and Alice and there is a Magic Door in a cave the magic door has a magic code that Alice knows but Bob does not so Alice wants to prove to Bob hey I know this code but she doesn't want to tell him he knows this code right so the process basically looks like you have path a and path B into a Magic Door that Bob can't see and so Alice will randomly choose a path and then Bob will say come back from path B and then she'll come back from path B and he's like come back from path a and she'll come back from path a so what you can think of this is if he asks her enough times and she can consistently come back in whatever direction that he wants her to come that means she was able to cross that magic door and therefore she also knows what the secret code is so that's interactive proofs when we talk about non-interactive proofs that's basically saying the prover can say hey I'm going to give you one proof or like one piece of information and the ver fire says that's all I need to know we don't need to interact past that right so the example here is about proving that anyone that I have completed a Sudoku puzzle this is a little bit more complicated so I'm going to dive in into it deeper but this is going to be what we think about when we talk about ZK rollup space snarks validity proofs Etc right so the way this works is actually something called hashing so hashing is a one-way transformation of data so that means like if I say oh I did 1 2 3 4 5 6 7 8 N9 I can hash it and give you some like long string of letters and you won't know the actual letter but what you do is you basically hash each cell's number in the Sudoku puzzle you combine all those hashes in commitments for rows columns grids um I guess I'm just assuming everyone knows the how sidoku is played um but maybe you don't so I'll step back again and also explain how sidoku is played in case you don't know cuz we're all about starting from the top essentially sidoku is this grid and what you need to do is be able to count 1 through nine from the top to bottom 1 through nine across the rows and one through nine in each box and there's no duplicates okay cool we all know what soku is if you take anything away from this talk I hope you remembered that specific piece um but essentially what you can do is create all these like commitments of the different forms of all the ways you created a zudoku puzzle and essentially the verifier is able to to use cryptographic algorithms to understand hey yes everything you sent me is every row you sent me fulfills the requirement every column you sent me fulfills the requirement etc etc but I don't actually know what the solution is right so that's how a non-interactive proof works so now let's actually talk about snarks versus Starks so maybe this is also something you've kind of heard and been like huh what okay um so we'll we'll dive into it right so snarks are succinct non-interactive arguments of knowledge so these are ZK proofs that are syn so they're small in size they're non-interactive so that means they don't interact with each other like the prover and verifier um these are very compact proofs that allow for fast verification thing to know is they do require a trusted setup so a trusted setup is a pre-processing phase that allows basically creators to create a public and secret key um for like I guess the actual proving the secret key needs to be thrown away so this is kind of I guess a con of ZK snarks um which we can talk about a little bit later in the next slide how do we solve for this so again another acronym that I heard and I was like what is this and I'll just explain it very quickly if you heard MPC you might hear it later this is called multi-party comp computation this is allowing basically a more secure way of creating those generating those keys right so rather than just like a single instance of like public key private key uh MPC essentially allows every participant to have a part of a private key and then work together without revealing their part of the private key to create the full one right so I mean there are trade-offs again right this doesn't scale really well when you have a lot of parties um it is vulnerable to collusion if like there's um I guess non- good non- good participants um but I'll I'll give another example so imagine all these frogs are pirates and imagine all the Pirates are looking for the secret treasure map um so they want to complete the treasure but they don't want to show each other what piece of the map they have so what they actually do is they divide each of their pieces into a bunch of smaller pieces and give everybody one part of their smaller piece so no one knows your complete picture of what your piece is but at the end of the day everyone does have every piece necessary to create the full map so that's NPC in a nutshell with frogs and Pirates okay moving on let's forget everything actually don't forget everything but now we're going to talk about snark Starks so Starks stand for a scalable transparent argument of knowledge so this is transparent because you don't require a trusted setup it's also non-interactive um this is really good for handling very large complex computations however the problem with them is they have larger proof sizes as compared to snarks and so this is where we thread again through data availability right the size of that proof is very important for when we post it back to the L1 because if it's bigger it's more expensive right um so I mean again these are different kind of ways that uh rollups have for move forward so like a starkware handle ZK Starks um however not All Is Lost um another interesting design space that I'm just tell telling you about that I learned right is there I guess math is cool because you can put math on top of math and so what that looks like in some solutions is you have a stark which is very good for computing large complex computations and then you wrap it you prove that it was true in a snark which is really good for making it smaller so an interesting thing to think about um but anyways at a glance you can kind of see uh snarks have a trusted setups Starks doesn't I put this here um postquantum means quantum quum resistant um basically snarks right now quantum computers aren't there yet so please sleep soundly but in the case that Quantum Computing gets there snarks aren't necessarily Quantum resistant Starks will be like I said um when it comes to developer tooling Starks are less I guess adopted so there might be less information in the space um when we talk about I guess a a ZK or a starkware for example they use Starks they have to use a different VM they have to use use different I guess they write in Cairo for example so super smart team it is a little bit different if you're like a solidity Dev moving in um and then I'm going to move forward hello yes uh no no so postquantum I should have used like the layman's term so postquantum means uh possibly Quantum resistant so what that means is quantum computers are not at a stage that they can hack stuff yet or hack snarks right but you know maybe we get into a world where that that is the way snarks are created is something called elliptic curve blah cryptography I'm not going to explain it because I'm going to butcher it but Starks use like random hashing and basically hashing is going to be Quantum resistant but U electric curve cryptography well that's a mouthful is not but we're not at the stage yet where we're thinking about it but but yeah cool good to know I will not write postquantum anymore cuz it is confusing thank you for that in the front okay another quiz time I put stretch sesh because I also expected people would stand up and be like I'm tired but I guess no one's tired uh oh nice thank you okay um what does snark stand for oh perfect okay I don't have enough swag for everyone who answered but if you find me first you will get a magnet okay moving on we have 30 minutes left so hopefully we can get through everything this is where we just get into current events um things I see on Twitter that people argue about that probably don't really matter in real life but whatever we're here talking about it so let's talk about rollups right I went over optimistic rollups I went over ZK rollups but did you know people Define more types of rollups let me list all of them for you that I saw on Twitter recently Sovereign enshrined based modular yes okay cool so let's actually just go through kind of I have like a slide for each we're going to Define it um and then you can find people who work at those protocols to explain it to you further but um before we start off we have to start from just kind of the base idea of modular blockchains right so modular versus monolithic monolithic blockchains do everything they do consensus they do settlement they do da and they do execution right so something like a salana would be a mod monolithic blockchain technically ethereum was a monolithic blockchain on until the rollup architecture was introduced right so modular blockchains do a subset of that so they might only do execution they might only do data availability they might do a subset in multiple um and so that's kind of when we get into all these new definitions of sovereign based enshrined I said something else I don't remember but we'll get into it okay cool so now we start with the basics again right so the what I talked about before you know rollups you might know today optimism scroll uh Stark net ZK sync Linea polygon Z KVM I'm sorry if I didn't name you and you work on that protocol um I respect you I just don't remember you right now um but essentially what those are are going to be what we call now smart contract rollups right so those are controlled by smart contract store on the L1 uh what they do is again like I said before store roll up state track deposits withdrawals handle verification key takeaway here is they mainly handle the execution part of it the other key takeaway here is something we're going to get into when we talk about enshrined rollups so enshrined Roll-Ups are different from Smart contract layer uh smart contract Roll-Ups in the sense that they integrate with the L1 at the consensus level so that means something has happened at ethereum core that they are now interacting with what that means specifically is when we talk about smart contract rollups right we have an external party determining what that smart contract looks like right we need security councils multi- sigs you need a bridge for up L2 upgrades um for um so when we talk about enshrined rollups you don't need all of that right because it's enshrined within ethereum itself um for a ZK evm enshrined rollup like validators don't have to re-execute everything because everything's enshrined in ethereum um the cons here are going to be VM in flexibility right so in this case you're married to whatever the L1 VM is uh the cool thing that I will talk about at the end of this presentation is about all the new different types of VMS that are coming about that's helping with scaling right so you lose that um the other is like it's slower to upgrade you don't move as fast um in terms of like development it might be more complex um but there are pros and cons like I said and there is a space where actually both of them live together um so we're all a happy family uh cool moving on oh I put a logo here this is tasos so tasos is an L1 blockchain if you don't know them they are doing enshrined rollups in some capacity I don't know how but find a tasos person and they will tell you um that's that's my uh tip for Devcon week uh the other piece you might have heard of is base rollups so base Roll-Ups is I guess not the traditional term of like you find an urban dictionary um but this is I guess the nerd sniping term for people who are interested in Tao right so base rollups basically use the L1 for sequencing so if you remember when I talked before about operators right operators are a centralized sequencer that kind of take everything together um base Bas rollups basically use ethereum for sequencing so what this allows ows us to do is inherit this decentralization of ethereum and inherits the liveness of the L1 so if ethereum goes down the sequencer goes down like uh we were we were Fu C ke D anyway versus like a centralized sequencer going down that's a different case right um and it is simpler when it comes to like decentralization of the sequencer um the con part of it right uh is uh if you want to get into it we can talk about it after I'm not going to go over it here but there are different performance improvements that rollups in uh do by like fixing the way they sequence stuff um that you obviously won't get in this case um and the other part uh uh maybe a good thing or a bad thing depending on who you are me that happens is going to be taken at the L1 level you no longer get like the sequence or MV me from l2s right because l1s are handling sequency the last piece we're going to talk about here is Sovereign Roll-Ups so Sovereign Roll-Ups are kind of like I choose everything myself because I am Sovereign uh so what that means is I choose my own execution I choose my own settlement layers um the Pros here basically are again talking about like that forced upgrade from Smart contract Roll-Ups right so smart contract rollup upgrades um if you want to participate in the L1 or if you don't want to participate in the L2 you just have to you have to exit right because the smart contract is upgraded sof and rups are actually like Hey we're going to choose our own settlement we're going to choose our own execution we're going to do everything our same it's what you can think of is the nodes control how they interact with each other so what that looks like is in order to upgrade a sovereign rollup you have to actually Fork like a tradition L1 um and so that allows us to basically kind of maintain whatever state we had before if you don't want to participate in the upgrade um the interesting thing here is right like this is a this is a design space where you might be using ethereum or Bitcoin as purely data availability and then not care about like um consensus settlement Etc maybe with somebody else right I didn't put a logo here but if you want to read more about it Celestia is a project that is working a lot on Sovereign rollups um but yeah so section four is like other Niche topics that aren't handled by rollups I yeah I don't know how much time I left but I think I have yes this is what I read online so the answer is yes as far as Google and a article that I read is correct um cool so let's dive into other Hot Topics so something you might have heard of recently is trusted execution environment so Ed execution environments is when we start talking about Hardware with regards to privacy and scaling right so a te is essentially an area within a computer's processor where you can run code but nobody else can access it so like the OS actually can't even access that space right um what this means it is isolated so it creates something called a secure Enclave is something you might have heard of within the main processor only pre-authorized code can run here um it has something not every te has this but some of them do they have attestations right so tees that you might use can produce an attestation to prove to others that the code is running is genuine and unmodified sorry I just burped and the last piece is confidentiality like I said because this is a separate um part of the I guess the hardware you get to do stuff that's encrypted and inaccessible to the rest of the system so something that's really important maybe for like privacy Solutions in that new space right um when we think about it further um actual te providers you might have heard of or now know because I am saying it right now until sgx is something it's going to be a hardware provider AWS Nitro is a te provider but they're more of a virtual machine um and like I said before this supports use cases like privacy it actually protects against me which is an really interesting space um flash Bots is doing stuffing in that direction if you want to Google Flash Bots te me um that would be a good place to start there's a question sorry what oh yes yes there's arm and trusso and I didn't list all of them there's a few but I only put two I should have put all of them thank you for that um but moving on uh why we care about tees right um it's a hardware secure environment um but to that point too it is going to be vulnerable to whatever Hardware SEC uh specific attacks there are you are putting trust in a different party such as an Intel such as an AWS which comes with its own tradeoffs um but these are very battle tested widely used um projects um it allows for fast computation with Hardware acceleration I think the con side is when we talk about like participating in networks if you require a te in order to run a e specialized Hardware it might prevent certain people from participating in the network um and then Integrity guarantees they have attestations um but again like I said it does require trust in the hardware manufacturers so whatever you have appetite for um but actually this is something interesting I want to dive into when we talk about a particular use case of tees when it comes to rollups so the L2 design space right and I mentioned before something called multi-prover so this is a new buzz word that you should keep in your mind so essentially if we look at systems today we have a basically a single proof system so if that is a if there's a bug in that proof system that's a vulnerability right so now there's a new thing that uh squirrel has done I believe Tao has as well if somebody else is doing please let me know um called multi proofers and what this essentially says is we want multiple proof implementations to basically reach a consensus of Truth in order to like strengthen the security guarantee right um there's different ways to do multile proofs right we have fraud proofs but that will introduce I guess you could say the challenge period so you have the problem finality running in again another way is just build a second Z KVM cuz you're crazy um so do it if you want that's I don't recommend it or maybe recommend it I don't know um and the third piece uh that scroll at least is moving forward with is doing a separate te right so we're doing the transaction in a te as well generating proofs there um and the way we're actually doing it again when we think about having multiple implementations um we are doing via different types of Hardware so if you see Intel sgx AMD AWS things like that um I guess another cool thing to think about when you're considering the rollup space uh what else do I have yes okay this is the last piece and I have 50 minutes left wow I talked for so long for so short of a time sorry our ZK evbm endgame so this is probably the spiciest thing right because I think youve probably been hearing a lot of rhetoric about like ZK ebms F fast finality like we want to be um I guess on par with you know the latest developments in ethereum but there's something that's really interesting in terms of the ZK VM space so ZK VM you can kind of think of It's relatively straightforward you just take the ethereum out of it it basically enables um like I said before computation that is provable but this time it's more General computation right so you're not limited to whatever the evm does um it's designed uh to support a broad range of applications the cool thing about ZK VMS that has allowed um developers to do um is actually build circuits in languages that are more mature right so if you're building circuits right now you might build in something called uh circum right when we talk about ZK EVMS that are bu being built right now we're writing custom circuits which requires Custom Security auditing which requires like I guess a very specific skill set but with ckvm what it allows you to do is actually do that same thing but do it in Rust right you have a bigger pool of developers um rust has its own like really cool like Speedy things that it does um and I guess projects to look out for that are doing ZK VMS is risk zero and succinct and this is really cool actually like I mentioned before of what this allows us to do um is actually uh succinct built something that allows optimistic rollups to start implementing validity proofs right it's it's very very new but this is kind of the new space that ZK VMS introduce right the other thing and like what that allows us to do right is accelerate ZK development in general so we think about blockers to privacy use cases ident Etc it's all like you you need a lot of I guess background information but now you can do stuff like rust and like I said before the auditing surface is easier um and so I guess new things to look up research zth and sp1 um ZK VMS are an interesting design space to follow okay so these are my sources that I kind of was like looking through if you want all of this information or want to read more um I had a pretty wordy slide because I feel like this is more useful as a reference but um this is a QR code to my beacon page which has a link to the the slides themselves so feel free to save it um if you want to follow me on Twitter feel free I mostly post about frogs so um I guess follow at your own Peril yeah I I think I have 47 minutes left so I kind of screwed myself over with 47 minutes of Q&A but uh that that's everything um I think that QR code is supposed to allow you to ask questions over chat but I'm also happy to just no questions yes no do do you write about real frogs do I write about real frogs um no I only post frog memes but I'll tell you about my favorite real frog CU I have 47 minutes to talk about my favorite real frog um and my favorite real frog is a desert rain frog which I have in a slide yeah uh [Music] uh cool this is my favorite frog and the cool thing about this is it never becomes a tadpole because it lives in the desert it just immediately becomes a little frog the end okay cool you seen the good place have I seen the good place I've seen parts of season one and then I fell asleep to seon there's a guy who really lik okay cool thank you if we're all good um you can also find me I guess at the scroll Booth where we'll have more swag for everybody who enthusiastically answered the questions so okay cool