Why L2 Makes More Sense Than L1 | Phuc Thai - Ronin

All right, I'm excited for our next talk on why L2s make more sense than L1's. Very excited to introduce Fukt Thai, a researcher at Ronin. Big round of applause for Fuk.

Hello everyone. Uh so my name is Fuk. I'm from Ronin. So uh a little bit of background on earth is that uh we currently an EVM L1 chain that focus on gaming and we are in the process of migrate into an EUM L2 which uh the mainet migration will happen in a month or so. Um so yeah today I'm going to talk about like why building an L2 is make more sense than L1 and um so it's the title could be a little bit controversial but uh so essentially what I going to talk about here today is sort of like a framework on how we assess uh whether or not we want to be an L1 or L2.

Uh so I think this could be uh helpful for people that are currently uh building an L1 a new L1 or considering to build a new chain. So the key consideration for this is uh there two consideration for uh for this is one is the security aspect of it and the second is the uh user experience. So uh a little bit of bricks analysis on uh the layer one and layer two is so for the layer one uh most of the current EVM layer one besides it uh it mainet is uh building on some sort of like proof stack. uh some of them be on like uh BFT style uh concessor protocol and it uh usually require uh on it majority to make sure like the securityities uh to make sure the security of network which usually is like it requires 75% of uh the validator who validating the network is on it um and compared with the itum layer two u which is essentially can inherited some of the security of Etherum. Uh so uh for for the state one layer two uh it only requires um on it minority which is usually uh the it depend on the setup of the security council of ex uh of the layer two but usually uh the requirement is that uh in order for security council to operate um the layer 2 it requires 75% of them uh to sign the signature.

So essentially it trans translate to uh as long as we have 7 76% of the circuit castle is on it then we can uh guarantee the security of layer two. Um but there's another aspect of the layer one is that usually um the valator on layer one uh had to st some of the token to uh be for them to get uh validating the network and there are some slashing mechanism in place. So uh in in a sense it can get uh economic security uh but essentially um yeah it's still um rely on it majority on the uh on the val set and when it come to the state two of the runup state two runup or state two layer two it can fully inheritate the uh the etherum security and uh there's still a certific capsule uh in place but there will be an exist window for the user when they can detect uh let's say the security uh council uh operates something uh bad uh and the user still had an access window to uh withdraw their asset before the uh the the uh upgrade actually take place. Um so and uh second aspect of security is on the cost. So on the for the layer one because uh it has the economic security and uh the valator had to stack the token uh before they validate.

So usually the validator incentive is essential for the validator to actually want to part participate in circulating the network uh which uh usually require like um certain emission emission of the token to uh for the circuit devices and usually uh when I look at some of the current L1 uh the annual staking zo is usually ranging from 3 to 10% uh of the staking supply. So an example on how much we had pay on the uh for security cost is like for our sale runin we had distributed approximately 100 million dollar on valid reward in the last uh three years. Uh so this this is actually a very uh significant amount uh when it come to uh the security cost and so that is the layer one security go and for the layer two security cost is uh mainly to uh pay on the settlement uh cost to it layer one uh which usually uh had the uh they need to post the state root on L1 the proof and uh the blob for data availability. So if we can look at the number here um so B chain is the over the last years uh BN only spend about 1.4 million on the uh uh the settlement cost uh to the EM1.

So and the number is significantly lower compared with uh the security cost of uh any layer one. Yeah. So another aspect of uh to to consider is on the user experiment uh when it compare between layer one and layer two. So uh because of the design of the uh layer two it inheritates the it inherit the security of it. The sequencer of player two can be centralized and because of that uh it generally um provide a faster block time and higher throughut uh compared with most L1 and the L1 can actually uh enhance uh to had a enhance the performance to have a high throughput and lower uh block time but it essentially it's straight off to the centralization uh for example right uh hyperlquid uh had uh also had fast uh block time and high throughput but then most of their val are also required to put in uh Tokyo which is also like introduce a cert trade up in the on this so in general uh the the the performance on the latency and the throughput of u the layer one is uh the layer two usually uh better compared with uh the layer uh so uh another aspect to consider is actually the interoperability when uh the liquidity that flow uh through different chain.

So one of the obser observation that we are see is like most of retailer they uh when they move the asset between chain they uh typically use a intentbased brit like uh relay lifi um and the canonical brit um that are mainly you for the fun transfer or you to rebalance of the intent bridge server. So uh for most of the time um the uh the the the UX uh when it come to crosschain using intentbased brid is not much different compared between uh the L1 and L2. Um but then there's one thing is that the uh the native rid of the layer two actually usually uh require a longer settlement time. For example, the uh uh if uh the base chain and the optimism chain like usually they has seven days of exist of challenge periods uh which is it take seven days to uh transfer a asset from layer two to it layer one and some other canonical bridge like CCIP or layer zero that's uh can settle this in in minute. So it is actually significantly uh faster.

Uh but then we also saw uh some of the layer 2 chain that actually just use a third party canon canon go brit alongside with some of the um with the native uh L2 brid. So essentially uh compare uh in the comparison of interoperability and the brid landscape of L1 and L2 I think it's not much different between the two. Um and uh the L1 also the L1 and L2 do not had much different but when uh I think it could change when we had uh sync composibility which is uh which also had been pushed by Vitalik and Eron foundation and uh when uh the sync composity and the base run up happened uh thing could change uh about it. So yeah in summaries um uh when we compare the layout uh whether or not we should choose to be an L1 or to be on L2 right. So from the security perspective uh we can say that um the the state one layer two actually in par in terms of security with the the existing L1.

Um and but then the and the state two like two actually had a better security uh uh is had lower risks compared with the other L1 and in in uh and also like the security cost of L1 is significantly higher than the the security cost of L2 and also the performance of uh L2 is also uh uh usually is better compared with uh the L1. So yeah, I think um this is also a framework when uh if people want to choose to to build a new chain either it's L1 or L2 it a good framework to to pick. Um so yeah uh that is is for my presentation. Thank you everyone to for listening.