Raising the Bar: A Deep Dive into Institutional Ethereum Staking Infrastructure

Transcript

Hello. Good to see everyone. Um, thank you for joining for this talk. Um, I don't have too many slides, so would love to go through the material and if anyone has any questions, happy to answer them as well. Um, >> [clears throat] >> uh, my name is Hong Kim. I'm the CEO and co-founder of Bitwise Asset Management. Um, and Bitwise is mostly known for our regulated financial products, uh, such as ETFs. Uh, we manage over 10 billion in assets. Uh, I've been doing it for over 8 years. Um, and uh, but what people don't know is that we, uh, are very active in Ethereum staking and also building the technology

involved, uh, in doing Ethereum staking, uh, in um, open source software contexts. So, today mostly we'll be talking about the technology stack and the open source software that we build and contribute, uh, for the space and also how that architecture of staking is different from, uh, maybe what you, uh, a typical home staker might do. So, this is the delta between maybe an institutional version of staking, um, what, uh, yeah, and and that's the focus of the conversation.

Okay. And maybe the main, uh, the ultimate institution, uh, in the Ethereum world is the Ethereum Foundation and the Ethereum Foundation started staking last year. Um, and uh so wanted to talk about the institutional kind of architecture and institutional staking from that lens of like how the Ethereum Foundation is staking, uh, and what they stated as their reasons of their decisions and then walking through those decisions, uh, I thought would be an interesting way of talking about, uh, institutional staking.

>> [clears throat] >> Um, and this was a tweet. So, they said, uh, this starting portion, uh, staking portion of this treasury, which is an, uh, landmark moment. And, uh, the next tweet is talking about the effort, uh, is being conducted with open source software options, Dirk and Vouch, uh, by Attestant. Um, Attestant is the company, uh, Ethereum staking-focused, uh, technology company that Bitwise acquired late 2024, so about a year and a half ago. Um, and uh, yeah, it talks about Dirk and Vouch, uh, and the reasons that those decisions were made as the kind of key components.

And [clears throat] that will be uh, the two um, open source technologies that I'll focus the talk on. Uh, there was that tweet, but there was actually kind of a following, um, uh, accompanying blog post by the Ethereum Foundation talking about the same subject and there's a little bit more detail, uh, so I I uh, uh, screenshotted the section there about the architecture decisions. And, uh, as they state there, after assessing many, uh, software options, the Foundation chose open source, Dirk and Vouch. Um, and Dirk is a distributed signer. The main thing is that it is remote distributed threshold

signer and it allows for, uh, the keys to be generated in a in a distributed way and also eliminates any single point of failure and adds redundancy. We'll talk more about that. And then Vouch is a multi-node validator uh that allows you to configure different consensus strategies and also provides kind of security and and and redundancy. And, [clears throat] uh, this setup allowed the Ethereum Foundation to, uh, employ many clients and also have a mixture of of, um, hosting context in, uh, a spreaded around in different geographies.

So, >> [clears throat] >> what does that mean? Uh, and we'll, uh, we'll go into a bit more detail here now. I'll talk about the Vouch first because it's a little bit easier, uh, to to talk about it first and then I'll talk about Dirk. Um, Vouch is a multi-node validator. What do I mean by that? A node, uh, you can think of as a single pair of a consensus node, a consensus client and an execution client. So, these are just examples of pairings that you could, for example, do, uh, Lighthouse and Geth, Nimbus and Besu, etc. Um, and what it allows is that if you were not using something like Vouch and, uh, just

running, uh, a node, um, in a validator, um, yourself, likely the setup you would have is that you would choose an execution client and a consensus client and run a specific pairing and you would run one node and that node is what, the node that your validator would interact with with to understand the state of the chain and also, uh, the state of the um, state machine, the execution environment as well. So, you, uh, your validator would engage with one specific uh, state of the world that is informed by one specific pairing of execution and consensus clients.

That is um, that is in most cases or in a in a in a kind of happy world, uh, when there aren't any client bugs, when we aren't in any kind of extraneous situations, is good and good enough for a lot of non-institutional staking contexts. But, in a context when where resilience or uptime, uh, or security matters more, then what is um, clearly, uh, beneficial is if you can actually harness the the what Ethereum affords. Ethereum affords client diversity, meaning that there are many,

uh, implementations of the same spec both on the consensus and execution side. And, uh, that is one of the real strengths of Ethereum that, uh, none of the other chains really have this to this to this extent. So, um, to you can actually leverage that to its full extent by actually running, um, multiple nodes in parallel. So, this this structure allows you to run, uh, four nodes in parallel. And what that affords is that you wouldn't stop attesting, for example, or proposing a block even if any of these clients have a bug.

And um uh, client bugs, uh, in in different kind of degrees of severity happen quite regularly. Uh, in a given year, maybe two or three times there would be, uh, some bug that leads to, it it might not be serious like an inflation bug or something, but more so, um, just something that still brings the client down. And, uh if and it's especially around hard forks and we're doing more hard forks these days because Ethereum is is progressing in a faster upgrade cycle and you often, maybe, um, in every other hard fork or so, there would be at least one client that that has some downtime, uh,

in the aftermath of a hard fork. Uh, so, yeah, this type of architecture provides protection against that. And is one of the ways that, uh, it's it's a core benefit that an architecture like Vouch allows for, um, and is one of the ways that an institutional setup of staking, uh, could be more robust than just the most simplistic version of it, which is running one node. And, uh, yeah, Vouch is an open source library. Anyone can use it. Actually, a lot of, uh, other institutional staking providers outside of Bitwise, use Vouch.

Um, uh, I believe something like 10 to 15% of all Ethereum blocks are proposed through Vouch. Uh, the run validators. Um, so, it's it's a meaningful part of the Ethereum, uh, consensus and block production. And, yeah, it's it's not just something that only Bitwise use and, uh, it's a meaningful we consider it a meaningful responsibility and contribution that we make to the space. And uh every every hard fork, we work with all of the client teams to make sure that, uh, the new versions of the clients work well with Vouch and, uh, go through every kind of testnet to, uh, leading up

to a hard fork to make sure that, um, the upgrade is is smooth seamless for all Vouch users. So, um, [clears throat] it's uh yeah, it's an important element. One more, maybe a little bit less known, the the the previous benefit is a more obvious and clear and everyone, I I think, or people that know about Vouch knows it. The, this aspect is maybe a little bit less, uh, known is that Vouch supports different types of strategies. And what do I mean by strategy? If in a world where you only have one node, consensus and client, um, pair, then you're just attesting to whatever that is saying is, uh, is true. But, if you have four

running in parallel, then you can actually decide what amount of confidence you would choose to actually attest to a block or propose a block. You can wait until there is In this case, it's looking at majority consensus. So, if you're running four that you could say, "I will wait until four of the consensus clients agree on what the source head and target of the chain is before I attest." And that would make the likelihood that your attestation is on the kind of the wrong side of future blockchain history to to be much more reduced

and and improves the overall network kind of consensus safety properties as well. So, in some ways, you're adding your node is adding to the security of Ethereum again in a way that harnesses client diversity more meaningfully. So, that's that's another thing that you you can do and this is was also actually a meaningful part of the Ethereum Foundation's decision to use vouch as well, I believe. Even from the if you look at the announcements before, there's vouch it talks about one of the main reasons as the variety of configurable

strategies that can be used to protect against client diversity risks. That is mostly talking about these strategies that you can use to fully harness the client client diversities powers. So, yeah, it's it's not what every institution thinks about, I would say. The kind of resilience or protection against client bugs, etc. That is a thing that is more intuitive and directly tied to consensus rewards and such. So, I think a lot of institutions think about that. This aspect is a little bit more nuanced, but the foundation obviously understands the nuance. So, seems to have been an important element of their decision to to to use vouch.

Um Now that we covered vouch, another thing to talk about is Dirk and um Dirk is a remote threshold um signer. What what that means kind of a distributed signer. What that means is that from the beginning there is a key ceremony, a distributed key generation event where there's never a single key that is held in one spot and it is generated in a way that the the kind of the shards of the key is

spread out across a set of machines of your choice. In this example, we're showing an example of three out of five. And in this type of setup, you get both better security and redundancy at kind of the same time because they actually are very related because the the way the Ethereum Foundation wrote it in the in the blog post is about kind of eliminate single point of failure. But if you think about eliminating single point of failure, it provides both security it eliminates single point of failure both from a key theft perspective or

someone kind of a rogue actor type of perspective and a resilient and redundancy perspective. So, in this case, uh vouch the the validator multi node validator will ask for signatures and to the Dirk instances that are at, let's say, different server locations or different bare metal server providers or cloud providers or or home hosted instances and it would collect three out of five signatures before attesting or proposing a block.

And even if two of the Dirk instances were down, and down can be just literally down because the the the server is down or it could even mean someone is malicious or even stole the key material and it would still be resilient against that both from a security perspective. One key shard wouldn't give you the ability to to to steal or sign signatures on the behalf of the validator. Nor would one shard one Dirk instance being down lead to attestations or block proposals or consensus duties being stopped in any way.

So, it removes a single point of failure in possibly the most important aspect of the validators architecture and operations. And so, [clears throat] this is a meaningful kind of I think the other popular library that most people use is web3 signer and in that type of context, even if yeah, you you would have kind of a single point of failure that if that instance was down, then you would have issues. So, that's what Dirk is about.

It's again an open source library, not something that just Bitwise uses. It's available it's been around both vouch and Dirk has been around since the inception of the beacon chain. It's been around for over 5 years and yeah, Dirk is also something that we contribute and maintain and think of as an important contribution that we make to the Ethereum staking community. And that's it. So, yeah, a pretty simple set of slides and this is the main way in which kind of the institutional version of the Ethereum staking is different from maybe a home staking setup.

Thank you. Any questions? I think we maybe have a few minutes. Any questions? You can also keep it short. All right. Okay, maybe that's it. Thank you.