The Execution Layer of EVM Chains: Scaling, Fragmentation | Ivan Paskar, Ihor Bielov | ETHWarsaw [4]

Today we would like to dive deeply into the kind of not an easy topic for all people who are just entering the web free but an interesting one for those who are already for years here. That's the execution layer of the EVM chains the scaling fragmentation and the future rollups. I'm Ior Bilof marketing social media labs and I'm here together with Ivan Pascar the head of growth at Altios.

Yeah, thank you. So I guess um you may be wondering why this is a big one. why why generally we should be excited about this one. So um I spent my background is like more than 10 years in crypto overall. Um I started as a web free marketer in 20 uh 16 um came through with a few agencies uh worked with hundreds of projects and then um I just finished uh my journey at Binance more than five years as a marketing lead there.

And uh people ask me like why do you choose to work in infrastructure uh project which is not sexy at all and not consumer oriented uh towards working at Binance. Uh so my response is that I generally believe in the multi-chain vision uh which we're going to see eventually and that's what we see right now like the recent announcement let's say yes from yesterday of what uh stripe is launching for instance um is a tempo blockchain and uh then we have more and more uh L1's coming out very soon. It gives me the feel that we are in the stage where it's not going to be just like hey use our chain it's the fastest one. Uh we need to come up with an infrastructure which will help to speed up and accelerate um absolutely all rollups and chains and it's going to be custom chain built by institutional enterprise clients whether it's going to be uh the existing cryptonatives. Uh so um a few points like generally why we should care um and I guess regardless of the fact that the buzz and everyone keeps talking about the prediction markets per uh stable coins um all of these going to be powered uh eventually by the new generation infrastructure uh which we have to take into account and um the generally like infrastructure has not been updated for years.

Uh at this point uh I remember um um part of my uh friends like they used to launch the decentralized exchanges in 2017 and it's been completely um irrelevant and impossible uh to work at that time by launching even the wallet and integrating it on Dex. It was super complicated. These days we can do lots of cool stuff. Uh well overall uh if we see that right now um like blockchains today they still struggle to meet the internet scale. What I mean by saying this is that uh if we would like to move onchain whatever uh is done offchain right now say could be social elements uh gaming um just like basic payments even uh what we can see is that 74 million uh transactions uh are run within a year only in US for instance only in uh like stock market and if we want this to be scalable then it says according to our analysis like you would need 2.

7K TPS I know TPS is a very blurry um benchmark but still we would prefer still more like the gigas benchmark by uh decoupling and understanding how many uh gas could be in the block size. So the question is always um what is the execution layer? Uh we've been pitched by this question many times. Uh apparently barely a few people know. Um so what we have right now in uh the blockchain system is uh generally there are a few standards right like a few pillars uh in um our like industry when it comes to the state access.

It's a data availability settlement execution uh consensus and yeah eager can tackle a bit about uh this uh by saying where we are and what role are we playing. Yeah and we if we are getting back talking in 2021 that most of people have been caring about the consensus and not really many people have been talking about the execution and as of now in the 2025 four years and people are still discussing data availability the settlement the consensus and we're still curious how is it going with the execution and uh that's the thing why we're here and why we're caring about the execution. ution and how it could be improved. the DA and the settlement are already actively explored but the execution remains the unsolved bottleneck and uh in case of the execution layer gap it's a great question how to scale and the real world examples and uh what can you get back into the web 2 and how it has been scaled for years at the beginning of 2000s the as infrastructure and how the web 2 apps have been exploring the opportunities to run on the new modern hardware so they can improve their TPS and uh they are metrics which are essentially for people who are using the apps and do not care what is on the back of those apps. So generally um if we are speaking of altis what we do for example this is something if we run like analogies in web 2 is kind of like radius or AWS uh for blockchains eventually we prioritize blockchains uh over protocols and apps because we believe that by doing this uh you can tackle more and uh what we do and why we focus specifically on execution side um is because consensus has been fixed uh I guess like for many years.

So we do have researchers and uh um really like techsavvy guys in the team who've done analysis and they uh say that there has not been proposed like any meaningful new consensus way although uh Celestia for instance they did a great job on the um data site uh availability and uh no one ever touched uh the execution um so it remains like the unsolved bottleneck uh and it can drastically speed up any chain which exists and that's what we are trying to tackle is like doing this multi-chain eventually of course we start with EVM uh which is the core backbone of um the whole industry and then moving to SVM OVM uh and so on by end of the year um I just realized this morning that um we are barely one of the um um like one of those uh projects uh here speaking at the war who uh does not has that much like recognition uh still and uh token as well. Um so I guess it still says something about uh how early we are overall in this infrastructure um uh case and the reason why we do this with the team we have is that um our team comes from the HFT industry. uh they are the very mature uh developers with a 20 20 plus years background. They come from the Oracle, Yahoo uh they've built applications like Shazam. So you definitely know that Shazam uh works instantly and that is kind of an example if we look at the consumer application what we are looking um to see in a few uh months years when there will be opportunity to build on top of u what we have and what we have is just a stack which is uh not a chain.

it would be way much easier, trust me, to just position and create a new chain. It feels like um this is like a wrong perception and there is no need to launch another chain. That's why uh we'd love to just uh help and assist any chains uh on these markets and if you are interested in um tackling this as well then we we can do and we can help with this uh because in uh the web two industry overall and high frequency trading uh the speed matters uh significantly. It's like nanoconds not even milliseconds and in our case in crypto we are far from uh these standards um yeah which is u why uh it is important for now

and in case of talking about the numbers we would like to mention about like finance the social and gaming are taking the most of the market and we can compare the blockchain with those already existing web two metrics for example the x generates It's 500,000 million posts per day with like three interactions each and uh 1.5 billion interactions per day. So it's 17,000 TPS needed and uh we have currently no blockchain that scales too much. Of course we're having like different blockchains and all of them are trying to solve the different kind of things which we're looking for in a blockchain. Some of them are focused on D5 and in case of D5 they are of course in need of the high TPS and some of them are the high frequency trade in and uh some of them are trying exploring gaming uh as some layer three solutions for example XAI which is on arbit room and they are building the game fight and uh in compared with Roblox which uh has 100,8 million people and like 17k TPS for one game which is still a big thing for web two and which we're aiming to bring to the web three and of course the finance that's why the most people are here like 2.

7k TPS and that's only the US market and uh it's globally more than 10k TPS and uh okay so the largest application uh types which are in need of that kind of execution is the trading payments consumer apps uh and of course the enterprise activity which we're seeing is growing rapidly in the last years and uh getting back to alu Ivan has already told a lot about altus and uh our main uh thing which we're focused on is the parallel execution and uh the memory first storage and uh as uh we've uh are aiming to scale Ethereum then the EVM has the largest base of developers and uh all of them are trying to scale Ethereum and to make it as an instant blockchain as mega ether are building super extra fast blockchain and uh with alt use we can bring that not only to the EVM and to the other EVMs in the future as well with the VM agnostic architecture and here is like an example of guess what we should know also like early bird gets the worm right so we can share with you that uh within this month we expect uh the benchmark to be released and our SSMT uh showcased the data that uh we are u tackling the gigas uh throughput latency uh which means that we're going to be overall faster 5 to 15x uh than average L2 like I mean the fastest L2 on the market and what it means is that we'll start working with them eventually uh so that they can improve their performance even better if it works then eventually of course we'd love to work with the Ethereum Foundation and um help Ethereum itself uh to be the fastest um in in its history. And you remember that Ethereum Foundation recently stated that uh they are looking for opportunities how they can uh reach 10K TBS. So that could be an option how we can do this. Um overall um the idea is uh just to give the competition opportunity to any chain and see what happens like if uh they can finally uh fix on what means like building some real applications and uh uh ideating like bringing that uh novelty from web two to web three. uh and because they won't be focused anymore on working and figuring out how do I crack the code and uh make my chain faster because we can do this uh together with them.

Uh then I guess the industry will win overall and um users will win eventually. Um what we see these days is like people are curious and uh wondering like hey okay like why should I care about you guys? uh feel feels like you are more for developers um you should talk to them. uh but eventually overall I guess we have not seen that a level of adoption and for instance if we even like tackle one um segments one industry in web 3 which is uh gaming no one cares about gaming right like we've even seen that VCs are um trying to invest trying to support but there is no progress overall uh in general within a course of five years and that's also the case because uh there is that's a level of um like infrastructure which can support uh what they can do offchain uh at these day. Uh but we are not just solely focusing on L1's and L2s.

So what we can do is uh maybe yeah um we can like give you a bit of perspective that we also help enterprises and institutions and that's one of the first partnership we going to be announcing this month as well is uh uh to build the custom chain um in the best way possible for like the enterprise institutional clients just like we see these trends happening these days with Robin Hood, Stripe, Google and so on

and also we cannot forget about reducing the gas posts for people who are just using the blockchain and uh using that on daily basis uh that's a big thing and uh for people who are validating on in the multiple blockchains especially on Ethereum like we remember Ethereum u two three years ago when the gas codes were like 500% more than we have now and of course the end users the all of us and uh one kind of memorable analogy how we can think of the execution layer in uh our everyday life and uh that's the formula one engine. So it doesn't matter which part in case of a blockchain we just plug it into the any blockchain and uh it can instantly uh like for the four weeks of uh the from helping from developer side from our team to the any blockchain in case of the first steps that CVM wants to uh go through all this process and to make them uh to enhance the execution of uh the current instead of the blockchain and uh it could be also the another analogy is like the ortor in the web two it's invisible but absolutely critical for the next failing everything I guess that's it for now if you have any questions please feel free to ask if not this is the website where you can find most of the information and um our team is mainly based in American's time zone but we are the only ones based in Europe so uh happy to chat offline. Um if you're looking for some um testing uh looking at the code on GitHub, uh it's right now in the beta stage uh before we roll it out publicly within Q3. Uh please talk to us and um yeah, I guess that's it. Thank you very much.

Questions? Yeah.

Yeah. Do we have any questions? Okay. Amazing. Hey uh do you have some examples of speed up which you already done in execution layer uh which you can show us some charts for example.

Yeah. So just yesterday we had a call with our developers team and uh they shared with us like the the benchmark data what what we can get eventually. Uh so it's not in production yet because we are pre-product stage but in a few weeks once we reload the product uh we can showcase. So the data says that uh as a minimum we have 450 meg uh speed. Uh what does it mean in a simple words is that an average L2 has maximum 100k meagas.

Uh which is like five x faster than an average L2 like base, BNB chain and so on. And um but the thing is that when we talk to these players like base and BNB chain and so on um one of the bottleneck and problem they see is also the memory. So there is um some specific number of uh block size and uh you can what they interested is to um input as much as possible gas in these uh block size so that more transactions can go in a parallel way. This is especially important when there is let's imagine I don't know like some NFT minting uh process or like claim process of a token when there is a very very high demand for uh simultaneous uh transactions uh interacting with one smart contract. Um so right now we see that the data also differs based on the blockchain.

Um they run obviously this data on the test nets and we would be curious to see how it works on the main net. Um the good thing is that uh we can do this without even requiring all validators to update their stack. Um so um the reason why we believe that maybe it would be more safe to start and see uh the real production and then showcase it to cryptonative teams is by helping uh to launch a chain from scratch to enterprise solutions. That's why our first partnership is going to be like uh institutional client uh in stable coin uh industry. Um very basic use case where you have transactions uh millions on a daily basis and um that would be the first moment I guess hopefully by end of this month when we can see the real put and example.

uh but on average what we are trying also to do is like to decrease the optimize the hardware use. So it means that um for example one of the best case right now what I see is like the real time of mega and uh what we can do is like optimize the 15% the hardware use uh and bring this to the VM agnostic uh vision so that we can do this on Salana chain although by the way Salana is already tackling execution stage by themselves internally uh but wouldn't say that like they are um at the final stage. Uh there is still some things we can help them to improve um even although they already started working on this. I hope that gives a bit of perspective what should be within like the next few minutes.

Thank you guys. Anyone else wants to ask something?

Hey guys, thank you for your presentation. So good question. A few weeks ago, I saw a bankless podcast with Vitalic Bhutan and he mentions that he sees the future for Ethereum to shift from EVM to risk VM. Are you going to be um compatible also with the risk VM as you mentioned? Are you that are you an chain agnostic for example or like um VM agnostic?

Yeah, I guess I can take this as well. Uh so when we saw this it's kind of like we had a very um unclear thoughts inside of the team because this is what we basically do right now and uh we realize okay that means that they are looking for this option uh but when we start talking to uh teams uh working with Ethereum Foundation as well then we realized that there needs to be some um checkpoint where you and showcase the real throughput and real examples. And ideally what we want to see is like some high network volume on these chains we've integrated with so that it would be like stress tested you know if stress test for exchange could be when they are hacked uh then for chains it could be when there is a high volume. So yeah, this is the long-term plan. Um, and uh, what is also part of our long-term plan is the open execution network uh, kind of like it's in the road map for next year.

It's a kind of like extension and improvement for our uh, stack. So what it means is like I guess you're all familiar with ka and cookie infoy stage. Uh we believe that developers needs to be also uh like rewarded and granted for contributing the good code and if they can optimize the code uh which going to be done in the way so that more transactions can be processed par in a parallel way then they will get rewards uh from altus uh as well and yeah hopefully this will uh kind of like um endorse them uh to write a better way and optim optimize because um when our core dev team looked at the generally all the code available uh on the market especially like EVM related they were shocked that it's not been optimized for years uh unfortunately um but it could be uh there is a way uh so it's going to be looked like available just very soon thank you thank you

okay great thank you so much