DVT and the Future of Ethereum Validators with Clayton Menzel, Jillian Freedman, and Brett Lee

Name of this panel is DVT and the future of Ethereum validators. So we have some lovely panelists here. Well, I'll go around and introduce ourselves. I guess I'll go first. My name is Jillian Freriedman.

I'm chief operating officer at Ether Capital. Ether Capital is a Torontobased public company. We're one of the largest ETH stakers on the public markets. We're staking about 80% of our 46,000 ETH treasury and we care deal deeply about a distributed and diversified Ethereum staking landscape. So on to you will.

Hi everyone, I'm Will. I'm a contributor to the Lido DAO uh in the NOM workstream. Very involved in all the work we're doing in DBT. So excited for this panel. Hi everyone.

Uh I'm Brett Lee. I lead growth at Oel Labs. Uh we are an R&D uh team focused on building uh infrastructure tech for Ethereum and other proof ofstake blockchains currently working on distributed validator technology and taking that to Ethereum mainet. Cool. And uh I'm Clayton.

Uh I work at Figment and Figment Capital. Uh we're an institutional focused uh staking infrastructure provider. Uh we also actively invest uh in the infrastructure layer on blockchain protocols. Okay. So let's get started.

The topic of this panel is DBT which stands for distributed validator technology. So our first question is what is distributed validator technology. And who wants to take a stab at answering that? I'll go ahead and and take that. Uh so DVT uh distributed validator technology essentially is a tech primitive that allows a single validator to be split into multiple nodes.

Uh this could be four different machines, seven machines, 10 machines. Each of these nodes can be run uh by the same operator, multiple operators. They can have the same client configurations, different client configurations. They can be located in the same geos or multiple geos. And the key value of this is to add uh is multi multiple uh layers.

But first is about resiliency of validators. Today, if you're running a validator on one machine, that machine goes down, guess what? You're offline, right? And the only way to add redundancy today is through what is called active passive redundancy. Which means you have a failover machine that has a full copy of the private key as a validator backup.

And if your primary machine goes down, you spin up the second one. Now there's a risk to that because if both those machines are up at the same time, guess what? That's a slashing event, right? And so there's not a great way to run highly available redundant validators today. Uh now in addition to that it creates a barrier to entry for a lot of solo atome validators.

Um and operators need to make sure they have great operations in order to become a solid validator. And so there's a barrier to entry for that. What DVD can do is to help solo validators as well now be able to spin up a node by themselves or with a group of friends uh four people, seven people, 10 people and not have to number one have the full 32 ETH to run their own validator but share that 32 youth amongst group of friends and on top of that even if again their living room machine goes down as long as their friends machines are still up that entire validator is active and and running. And so just to finish up this uh this narrative here uh we recently launched actually the first ever DV on Ethereum mainet core team did it uh all living rooms uh four houses in three different countries in Ireland, Estonia and in Canada and uh the distributed validator is performing at the top 1% of all validators on uh the network. So it goes to show that you can use this technology to make at home validation much more accessible.

Thanks. Um, so just zooming out a little bit and looking at like why why is this important? Um, will and anybody else that wants to add also, what what role can DBT play in the the decentralization but like the overall resilience of of Ethereum? So I like to sort of think of this from the micro and the macro. And the micro being the individual validator level uh flowing up to the macro which is the overall Ethereum network.

And so today you have a single validator that has an execution layer client, a consensus layer client. Um and DVT allows you to take both of those things uh that that single validator split it across four, seven, 10 and combine different combinations. On top of that you might have a validator running out of someone's house in Canada. Now you can have a node in someone's house in Canada, uh an AWS data center in Europe, uh another house in Australia and another house in Mongolia. Um so on top of this, you can combine different kinds of infrastructure types.

So different hosting, different locations, different clients. While that happens at the single validator level, you've made this thing so much more resilient to single points of failure. Those risks uh become much less. And when you adopt this networkwide, the overall network becomes much much more resilient. So it also sounds like DVT is a gateway for smaller operators to get involved.

Um I'm wondering Clayton if you have any insights in like what is the role that like larger node operators like a Figment would play in the buildout of DBT? Yeah, definitely. So, um, if you're someone that doesn't want to run your own infrastructure, um, for compliance reasons or because you're not, you know, just don't want to learn how to do it yourself, um, there is a, um, you know, single point of failure for any sort of client that you're using or if you want to work with multiple providers, um, you know, those providers themselves, you know, are managing, uh, their infrastructure internally, maybe using proprietary sort of solutions. uh DBT allows you to work with multiple different node operators um and allows you to you know have a lot of visibility uh in how it's functioning um you know as as a you know token holder wants to stake uh without managing their own infrastructure um I think also because it allows you know these validators to be um more resilient right like you are having you know high higher effectiveness rating higher rewards uh less risk from a slashing perspective uh it just you on the adoption curve of whether being a node operator or a staker or someone that's interested in the idea of staking Ethereum just becomes a lot broader um you know down to the um you know at home you know hobbyist Ethereum community but like also to larger you know asset managers um like actual institutions as opposed to uh you know crypto institutions. Um so something that comes to mind is MEV and how does MEV work in a DVT world.

So in short uh this is kind of in the research phase. Um the way uh we think about it is like they're both part of a new kind of burgeoning middleware layer in Ethereum and it'll be really interesting to see how distributed validators DBT technology and MEB boost and other middleares that will come up in the future kind of interact together. Uh but yeah so more more to be more to be seen um I think and there's a lot of research to be done in how these different middleware technologies start to fit together. Can I one thing I I I think that's interesting about the like me topic in DBT latency is such a huge problem for me and it's like one of the potential issues for DBT and so I I think that's going to be when we talk about adoption of DBT probably one of the things that that you know is a bit complicated as we like configure setups because even today the merge was what like four or five months ago we still have problems with uh me boost like some days with latency. So no I I agree and I think that's definitely a consideration.

So we're kind of in our second uh test net right now. It's called BIA. And one of the things we're focused on is making sure that the performance doesn't degrade even as you're distributing these validators. And so so far contrary to I think what a lot of people expected, we're actually seeing inclusion delays that are that are lower than what you would expect. So uh we'll continue to monitor that.

But yeah, absolutely performance latency is a critical thing to maintain even as you start to decentralize and distribute these validators. Is is there like a sweet spot of number of participants in like the example you said where you guys launched something on mainet and you had four different like friends running OBL setup is there like what's the range of unique participants? Yeah. So right now I mean the way we think about it the ideal setups are four, seven or 10 participants and the reason for that is it optimizes for the security of uh the cluster. When you have a four-person node, uh you can have one node go down and that cluster is still actively validating.

When you have seven, you can have five. When you have 10, you can have uh seven uh active and still validating. Anything in between that, you still need to meet those thresholds, which means you need a higher percentage of the nodes to be active in order for that validator to be running. And so, basically, 4, 7, and 10 are the ideal setups that we've seen. We haven't uh really tested any more than 10 cuz you don't really need to.

Um you could have a lot of validators running on these distributed clusters. Uh so but that's that's kind of what we've seen uh so far. Um, and I want to ask a question about liquid staking, but I have one follow-up question, which is, do you see the need for some sort of a like independent community of unrelated people that would like pair up to actually run these things and how does that look and what is the what is the future of that? Yeah, so it's funny that you asked that. the the first version of Opel is going to be kind of a trust model.

And what do I mean by that? You kind of need to know who you're running your validator with, right? With that said, it doesn't have to be somebody like next to you. You can find this person uh wherever, right? And so we're running again our B test net right now and we have over uh 900 clusters made with people around the world purely finding each other on a discord channel.

Like they literally said, I want to run a validator. Do you want to do it with me? Great. Let's go. Right.

And sharing a little bit about a validator dating app. Exactly. Exactly. So you can definitely do that and in the future in V2 we are uh looking at more trustless models where uh you can join and not know who uh your other you know partners are in the cluster and still be protected uh from bad behavior, visit team behavior, things like that. But I I also think DBT is one of the missing links for this sort of uh open marketplace for you know collecting stakers and uh node operators because of you know the preference associated with clients or or geographic locations.

Um you know that's not necessarily a a necessity for someone that may be building a liquid staking protocol. Uh but it is obviously something that like gives people a wider variety of risk and reward appetite as well. Um, will I wanted to ask what are some of the opportunities and challenges when it comes to liquid staking and DVT? Yeah. So, I'll try and contain my excitement because I think DVT on the uh on the liquid staking side, especially 2023 is is going to be huge.

Um, currently Lido, I'll speak from Lido's perspective. We're we're running our second pilot on testnet uh with both Oble and SSV which is the other uh sort of large DBT solution that exists today. Um, Lido announced the V2 uh upgrade a couple weeks ago and basically this allows for the modularization of the protocol and what that really means is that we're going to allow different kinds of node operators to start participating in different ways and DVT is going to be a core part of taking Lido from 29 Ethereum node operators today to hopefully you know 100 in the next 12 to 18 months and very very soon after thousands. Um, DBT allows basically us to combine solo stakers, community stakers, professional stakers like Figment and create really resilient decentralized clusters that allows Lido to, you know, decentralize our node operator base to a huge extent. I think coordination is going to be one of the issues still in 2023.

uh DKG or distributed key generation uh is something that we'd love to see in an ideal world that still requires a fair bit of manual work like the golden dream like years down the road would be you know you have a solo staker a community staker professional staker all ready to be allocated ETH and DKG would happen asynchronously and the cluster would basically spin up and no one would touch it I think we're still a few years away from that but DBT is going to be huge for Lido and liquid staking in general this year Um, thank you. And I I I guess like we're we're still very much in earlyish days of this and so there's still a lot of questions. Um, Clayton, I wanted to give you an opportunity to just add in anything that might have been missed in terms of how the staking community might evolve and like what the impact of DBT will have on the staking community. Yeah. I I mean I I think about this from like you know the an adoption curve sort of perspective and you know what came before it uh to where you know if you think of you know proof of work versus proof of stake adding a new validator to the Ethereum set is like adding a new minor to the Bitcoin network.

Um and so solutions beforehand like you know Lido or MEV is just incentivizing more stake on the network. um a solution like DBT um actually makes that stake more valuable uh because of how efficient these uh you know nodes can become. So I I think that's very interesting uh where you know it's uh I mean it would hard to be like gauge an actual number uh but like how much valuable that the network would be uh in relation to how much Ethereum is currently staked on the network because of it operating that way uh to where it's not just a benefit from a staker perspective anymore. It's a benefit to everyone that's interacting with the execution layer as well. um which I think is highly interesting.

Um I think DBT bleeds into resaking which is another thing associated with being able to export the value of Ethereum security to other ecosystems. Um and that security is more valuable because of DBT, but it's definitely in the very very early stages. Um and we'll see what happens. Um, yeah, super interesting. And I know we were talking about that before and how like DVTN restaking works and it sounds like we're it's really like the bleeding edge of the bleeding edge.

Um, so a lot of questions remain to be answered. Um, I wanted I see we only have a few minutes left and and maybe there are questions from the audience, but I wanted to put out this more like uh existential question which is like what in a world without DVT can Ethereum remain credibly neutral and what doesn't like what is the alternative if like no DVT DVT dies on the vine like knock on wood like that won't happen but like why is it so important. What happens to Ethereum without a DVT like solution? I I mean I think without uh you know it if you can't incentivize diversity within the validator set uh I a the best solution on the market will have concentration right um and so over time without being able to merge everything together through something like distributed validator technology you're going to have a centralization risk same with like mining pools on Bitcoin same with mining pools on Ethereum less you know, a few months or a few months ago at this point. Um, and you know, just concentration at the at the staking level as well, right?

Like there probably would be a handful of actors that would be able to perform the function better than others. Uh, and the market will just move over to that. Same with any other market that exists um on this planet. Yeah, just to add on to that, I think just without DBT, there's a lot more risk in the consensus of the network. uh whether it's centralization risk, whether it's uh slashing risk, uh whether it's offline risk, you know, there's just a lot more risk without this primitive embedded into the protocol or into the network um to kind of decentralize and to make the entire network more resilient.

So, I would sum it down to to risk, higher risk. Okay. Well, that's super interesting. Um I'm wondering if there are any questions for the audience. If not, I can keep lobbing a couple at you guys.

Yeah, it's a it's a great question. So, actually we launched a product called the uh distributed validator launchpad or DB launchpad. Um, our head of product actually was the one that built the Ethereum or ETH2 launchpad uh back in 2020 and it basically provides a a user interface for you to go in uh you know and then configure the cluster and then run through the steps to be able to activate it. So that's how all of our uh test net participants you know at home validators are spinning up their clusters is through this UI interface and it basically walks them through the entire process of spinning up a cluster. So you need to get your machine you know set up the machine and then run through this uh DB launchpad process and then you have a you know distributed validator setup at the end of it.

Any other questions? Okay. Um anyone have any final words to add in the last 30 seconds? Uh I would just say would like all of you to get involved one way or another. Like if you are a validator yourself uh prof part of a professional entity come talk to us.

If you're a solo validator love for you to participate in you know testing the technology and getting it to mainet and then it also if you're not if you're just staking ETH I think just to be more familiar with DBT is important because it is an important security you know primitive and security standard if you will that needs to be implemented into the network. So I think just getting more familiar with DBT is what I would ask. But yeah, happy to talk after this as well. I'll throw in two quickly, but I I think uh Lido will definitely be doing more rounds of testing with DBT, solo stakers, community stakers, other node operators. Please reach out.

Uh we'd love to have you join in our future pilots. Figment will be using DBT as well. uh and also the the importance of you know open source solutions like this that are universally accessible not you know doubling efforts at every part of the stack where everyone has to build a proprietary solution is not the direction of this ecosystem moving forward and so we're we're very excited to you know be working with with both teams awesome well thanks a lot guys we're over time but thanks for everybody here for listening thank you and I guess we'll stick around if anybody wants to chat uh afterwards hi Hey.