Overview of DA solutions landscape - Maciej Zygmunt

Transcript

[Music] So yeah uh hi everyone my name is Machek I work at L2bit as a software developer and today I want to give you a quick overview of data availability landscape. I will also tell you how blobs/ rollup fees and why teams are now exploring alternative DA layers plus the trade-offs you should keep in mind. So let's dive in. EAP 4844 gave us blobs, a special space where rollups can post data for a fraction of what call data cost. And we saw a drastic drop in rollup fees from a very first day as most of the rollups switched to blobs almost immediately. Check the charts.

First base far higher volume but much cheaper transactions. Activity went up from few user operations per second to more than 100. But transaction cost went down from half a dollar or even more to fraction of 1 cent. RB room, same story. More transactions, but cheaper transactions. And many other rollups follow the same pattern. So yeah, it's all great, but the blob space is full almost. The original blob target was three blobs per block with a hard cap of six, and we have been hitting that sailing since October last year.

Pectra upgrade in the first week of this month rise the target to six and a hard cup to nine. And even though we haven't maxed it out yet, uh the average blob count is climbing every week. So we can expect that it will be filled soon again. Also, uh it even though it's not full yet, there's already no room left for another big project like Bass as B itself is posting 44% of the whole data posted in last few days. Uh and with new rollup popping up almost daily, saturation is just a matter of time. Also uh whenever the blob count goes above the target the blobs base fee is rising and blobs are becoming to be expensive. Before the uh

before the pectra it caused a lot of problems to some rollups when bass was pushing a significant amount of data. Some rollups like optimism arbitum and all the big ones just switch back to posting to all to call data which is good but some really small rollups just stopped posting completely what which was hurting their livveness. Currently this is not the case but we can expect to get there soon again. So we won't be able to push more data and go higher with L2 throughputs without sacrifying livveness which means how frequently rollups post data to Ethereum while theorically rollup could gather all this transaction data compress it and post it once a day. It wouldn't provide the user experience we desire. When I'm doing a transaction on L2 I wanted to be

included in L1 block as fast as possible to have the finality quickly. So yeah because all the all these constraints a lot of high TPS projects like Eclipse Mega and many more lead real-time blockchains are now exploring alternative DA layers. So yeah enter altas chip and roomy. First up is Celestia modular DA chain. Yesterday it handled a bit less than 3 GB of data and yet it can handle up to 120 GB of data. So only 2 and a half% of space utilization where Ethereum target is at around 5.3 GB of data daily. Next aail on only less than uh 1% of space utilization with a max of 17 GB of data

daily and many other LDAS follows the same pattern. Again with throughput of 15 megabytes per second, espresso DA on test net with 30 and near DA with more than 53 megabytes per second where Ethereum is at less than 0.1. So yeah, more place for still very cheap competitive price. It feels perfect, right? No, because your data now lives outside of Ethereum. So new risks appear. And to gather all those risk, we at Lubbit created a dedicated data risk framework. It covers in total five different risks split in two buckets.

First to cover DA layer risk and other three DA bridge risks. DA bridge risks only apply if project is using DA bridge. Skipping the bridge is even worse as Ethereum has no proofs that the data was ever posted there. And the A bridge basically acts like an oracle that brings back the commitments back to Ethereum. Unfortunately, currently there are only few projects that are using uh the A bridge. So yeah, let's cover the whole framework quickly. First risk from the layer risk is economic security which measures how much stake the validator the a validator would need to burn in order to perform a successful data withholding attack. And the green risk applies when bra risk is measurable onchain plus uh slashable funds exceed

total value secured by this de layer. So basically the slashable column on our website should be bigger than total value secured column. Next risk is fraud detection mechanism. So which assesses like how effectively users can protect themselves if majority of the validator or committee members turn their malicious and this is basically data valability sampling check. Next risks from the abridge risk committee security which evaluates the committee attesting to the DA bridge. Uh this scoring is really very similar to our security council scoring in our stages framework. So basically how independent and uh diversified are the members of the committee. Next one upgradeability. This one is very simple.

So if there is enough time for user to exit in case of unwanted or malicious upgrade. And the last one relayer failure. So what would happen if relayer fails or misbehaves and the green risk applies when user can self-propose DA commitments because if is not the case then relayer becames the livveness bottleneck. So yeah to sum up everything very quickly in current state of DA blobs Ethereum blobs allas can be a good solution but not for every use case given the new risk that we have just covered especially when not using DA bridge we need to scale Ethereum blocks more so projects don't even need to consider choosing aldas and there will always use Ethereum DA as a default

fortunately we already have the another step on our road map the Fusaka upgrade which should take place by the end of the year and should scale Ethereum blobs from a target of six to to 48. So 8x improvement. The goal is straightforward. Make Ethereum blobs big enough, cheap enough to handle all those rollups data securely, affordably here at home on Ethereum. Thank you. [Applause] So thanks a lot Machek for your wonderful presentation and if you guys have any quick question please feel free to ask. We don't have much time. So we will take just one question.

Please don't be shy. Okay. So I will ask you a quick question. So what's your next milestone that you are planning to achieve? Uh like in Ethereum DA? Yeah. Yeah. Yeah. for sure the fusaka upgrade which should raise raise the blobs for 10 8x and uh after that we will know we will not know now we will see how what will be the the the dependent for Ethereum blobs yeah okay we have question can you go again to the blob space diagram is like I would like to know why there

is like um it doesn't work. Uh but do you remember like the one that you were showing with base and all the layer 2 like there was one period without any like blob space like for unit chain and and also like base was like a bit bigger in that moment. Do you know why uh what happened there? But wait, which chart? The with the gas fees or no like the blob space I see. Ah yeah. So why the base is is bigger? Yeah. No, not the base. But I mean there was one moment uh which like zero blob space like yeah like we can uh yeah so like I I'm not sure at which which which part of the chart you are

you're talking about but maybe we can catch me after the the the presentation and we can talk about it. [Music]