FCAT Crypto Brief™

Tim Garcia discusses Solana's Alpenglow upgrade, governance, and validator diversity at Breakpoint 2026

30 min Listen to the original Episode page
SolanaBlockchain InfrastructureConsensus MechanismsGovernanceValidator Relations

Summary

Tim Garcia, Solana Foundation's Validator Relations Lead, outlines Solana's goal to be the fastest blockchain for on-chain assets. He details the Alpenglow upgrade, which reduces finality to 150ms and block times to 200ms by moving voting off-chain. Garcia also explains the evolution of Solana Governance Proposals (SGPs), noting SGP 002 passed via delegator overrides. The conversation highlights the importance of client diversity, citing Fire Dancer and Mithril, and emphasizes infrastructure resilience over simple validator count.

Key takeaways

  1. 1
    Solana Governance Proposals (SGPs) allow nuanced voting and delegator overrides, built by the Solana Foundation, Exotech, Turbine, and others.

    Unidentified speaker (speaker_0) · 5:48

  2. 2
    AlpenGlow is live today, the Alpine Alpen Glow consensus layer. And it allows for about 150 milliseconds in finality. So that's on the level of like web two clients or web two applications, right? Where you can go click a button and then it's done. It's already finalized.

    Unidentified speaker (speaker_0) · 9:50

Episode sections

  1. 0:00 – 5:11

    Introduction and Validator Role

    Tim Garcia explains his role in coordinating with validators, infrastructure providers, and core developers to ensure network uptime and protocol evolution. He emphasizes that Solana aims to be the fastest and cheapest blockchain for on-chain assets and traders.

  2. 5:11 – 9:16

    Solana Governance Proposals (SGPs)

    The discussion covers the evolution of Solana governance from a clunky on-chain system to SGPs, which allow nuanced voting (yes/no/abstain) and delegator overrides. SGP 002 (double disinflation) passed with delegator overrides, and SGP 001 addresses confusion around abstain votes to improve the voting denominator logic.

  3. 9:16 – 18:24

    Alpenglow Upgrade and Performance

    Alpenglow is described as the biggest protocol change in Solana's history, reducing finality to 150ms and block times to 200ms. It moves voting off-chain to reduce block size, freeing up capacity for user transactions, with further optimizations like 'Rotor' and multi-concurrent proposers planned for the future.

  4. 18:24 – 29:37

    Validator Diversity and Network Resilience

    The speakers discuss the importance of client diversity, noting Fire Dancer has over 16% stake and the Mithril client has launched on mainnet. They highlight that security relies on diverse infrastructure, geographic distribution, and supply chains, rather than just validator count, with Alpenglow adjusting the finality threshold to 60%.

Assets discussed

AssetKindStanceContext
SolanaSOLcryptoneutralI'm very excited to be reporting today from the Solana Breakpoint Conference here in London.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Solana transaction end-to-end wait time is about 350 milliseconds.

Unidentified speaker (speaker_0) ·0:00

How to verify: Network performance metrics from Solana block explorers or official Solana Foundation documentation on latency. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Solana has a six-week release cycle.

Unidentified speaker (speaker_0) ·2:52

How to verify: Solana Foundation release notes or GitHub commit history for Agave/Anza clients. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

SGP 002 (double disinflation) passed by a thin margin due to delegator overrides.

Unidentified speaker (speaker_0) ·6:28

How to verify: On-chain governance records for SGP 002 showing vote counts and delegator override statistics. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Alpenglow provides about 150 milliseconds in finality.

Unidentified speaker (speaker_0) ·9:50

How to verify: Solana network performance dashboards or Alpenglow upgrade documentation. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Current Solana block times are 200 milliseconds.

Unidentified speaker (speaker_0) ·11:28

How to verify: Solana block explorer data calculating average slot time. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Alpenglow moves votes off-chain, reducing block size.

Unidentified speaker (speaker_0) ·13:57

How to verify: Technical documentation for Alpenglow consensus changes and block size analysis. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Solana has 100 million CUs (Compute Units) available.

Unidentified speaker (speaker_0) ·14:17

How to verify: Solana runtime documentation or programmatic checks of block capacity limits. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Fire Dancer has over 16% of stake.

Unidentified speaker (speaker_0) ·19:47

How to verify: Solana stake distribution data from block explorers or stake pool analytics. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Mithril client is on mainnet running on a validator.

Unidentified speaker (speaker_0) ·20:07

How to verify: Solana validator list or Mithril client announcement confirming mainnet deployment. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Alpenglow changes the finality threshold to 60% from 33%.

Unidentified speaker (speaker_0) ·23:12

How to verify: Alpenglow upgrade specification or Solana runtime source code regarding consensus thresholds. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify Solana Foundation's official blog posts regarding the Alpenglow upgrade specifications and release timeline.
  • Check Solana's on-chain governance records for SGP 002 voting results and delegator override statistics.
  • Review Fire Dancer and Mithril client documentation to assess current stake distribution and mainnet readiness.
  • Examine Solana's developer documentation for the 100 million CU capacity metrics and off-chain voting implementation details.
  • Consult Solana Foundation's validator relations guidelines to understand the six-week release cycle and patching requirements.
  • Analyze third-party blockchain performance benchmarks comparing Solana's 150ms finality to other Layer 1 networks.
  • Review Solana's GitHub repositories for Alpenglow and Rotor upgrade code changes and community feedback.
Full transcriptexpand

0:00Unidentified speaker (speaker_0): So end to end, let's say if you send your transaction at the very beginning of a block, the total wait time would be about 350 milliseconds, which is, you know, quite fast, but we're trying to push that even lower. We have thoughts and discussions about making the block times even lower, like 100 milliseconds, 50 milliseconds.

0:20Unidentified speaker (UNKNOWN): Get

0:20Unidentified speaker (speaker_1): ready to jumpstart your week with the FCAT Crypto Brief, brought to you by the Fidelity Center for Apply Applied Technology. Join us as we delve into the hottest topics, from blockchain breakthroughs to DeFi revolutions and market happenings. We promise go deeper than the headlines and deliver market insights you can't get just anywhere. Be sure to hit subscribe for your bi-weekly dose of crypto commentary. Quick disclaimer: we're here for education, not investment advice. Use Views Express. expressed are of the hosts, not Fidelity or its affiliates. Crypto's a wild space, highly volatile, can become illiquid anytime, and is strictly for those with a high risk tolerance. Buckle up, enjoy the ride, and let's dive into what's been happening recently.

1:04Unidentified speaker (speaker_2): Welcome to the Crypto Brief. I'm your host, Jason Ward. I'm very excited to be reporting today from the Solana Breakpoint Conference here in London. We're joined by Kevin Kelly from the AFCAT Validator team and Tim. Garcia, who is the Solana Foundation's validator community lead. So very excited to be here. There's a lot of energy. This place is huge. There's been a lot of pre-events going on, but from the FCAT perspective, we're here to learn and actually talk a little bit more about what we're doing in the community. So, Tim, it's very great to have you with us. There's a lot going on. So really appreciate you being able to take time out and join us. Just maybe for those who aren't familiar with you, what is the actual role of the validator community lead and how did you come into crypto?

1:42Unidentified speaker (speaker_0): My role has varied over the years, I would say. But generally, my job is to talk to validators, talk to the infra layer, talk to bare metal hardware providers, and make sure everyone's on the same page, make sure everyone's updating on time and work with the core developers as well to figure out what the next big things for the protocol are and how we move Solana forward. A lot of our jobs is how do we make Solana the best place for assets and for traders on chain?

2:09Unidentified speaker (speaker_2): Excellent.

2:10Unidentified speaker (speaker_3): Yeah, Tim, I you know, I think it's understated that validators really are the backbone of any network, any blockchain, and critical to uptime blocks getting appended, added to the network. Expanding on Jason's question, how would you also describe? the role of validators in in security and maintaining liveliness for the network? Yeah,

2:32Unidentified speaker (speaker_0): you know, not to brag, but I I think that Solana as a community has the best validator set in the industry. I think a lot of what the validators do are, you know, stay really involved in the development of the protocol. We ask a lot of our validators, so they have to update frequently, I'm sure as FCAT knows, running a validator yourself. You have to stay patched. We have, I think, the fastest pace of development in the industry as well. So we have about a six-week release cycle, which just requires a lot of vigilance, a lot of involvement, a lot of activity from the validator set.

3:05Unidentified speaker (speaker_3): Yeah, it's no joke. It's, you know, once an upgrade a week, right? At least. Yeah.

3:09Unidentified speaker (UNKNOWN): Yeah.

3:09Unidentified speaker (speaker_2): When you say a six-week development cycle, that's insanely fast. So, you know, with all the these different operators around the globe, there are different channels that you're using to keep people informed and keep people synchronized on what's happening. But what does that actually entail? Are you hosting community events, monthly updates? How do you get that information out to the validator set? Yeah,

3:29Unidentified speaker (speaker_0): a couple different ways. So we have a monthly call that the Solana Foundation hosts. It used to be twice a month, and I was running those calls, and then we distributed a little bit more. So the there's a few validators in the community, ChainFlow and H2O nodes. They've kind of taken that mantle and they have a community-led validator call. I've joined it as well, but there's a lot of discussions there on what's coming next with the protocol, what do you need to look out for, updates from the development teams, Anza and Fire Dancer, for what's coming next. So that's a big conduit for information. The Discord as well is is where we distribute. a lot. So the validators all kind of hang out there and chat there. There's a lot of arguing and you know discussion that goes on there. And then there's email blasts, there's slana.com slash upgrades where we have a lot of information as well. So just as many channels as we can to keep validators updated and informed. That's

4:21Unidentified speaker (speaker_2): great. I'm sure the validator community is pretty lively there. Uh Kevin, I think you're spending a ton of your time on Discord and what's it like for you as an institutional validator operator? How are you taking advantage of what Tim and his team are putting out there and just general community um outreach?

4:35Unidentified speaker (speaker_3): Yeah, definitely. We are, you know, I think as with any other validator operator, you do have to keep a very close eye on GitHub releases for the different clients. You know, Tim mentioned the Discord channel. We keep a very close eye there. Is that as that's the source of news and updates for high to low priority upgrades. So yeah, it's as an operator, you have to stay very vigilant. And sometimes there are patches you have to put out within 24 hours, you know, to keep up pace with the wider validator community and to make sure that you're supporting the wider network. I think the something we haven't mentioned yet are the Solana governance. proposals, SGPs that I think are a little newer, right, Tim? And yeah, curious, you know, what's the state of SGPs at the moment and why should maybe our audience keep an eye on those going forward?

5:28Unidentified speaker (speaker_0): Yeah, maybe a little bit of background is useful there. So Solana governance has evolved over time. Uh an SGP stands for Solana Governance Proposal. And Kevin was saying that's a relatively new thing. So in the past, we've had an on-chain protocol for doing governance, but it was a little clunky. Validators didn't really like it very much. You couldn't really say anything except for voting yes. There was no nuance between no and abstain. And they took it upon themselves to build a different governance system. uh, But that still was a little hard to manage. So the Solana Foundation, along with Exotech, Turbine, Jito, and some other players built tooling on-chain to support a more robust expression of validators saying yes, no, abstain. And actually, one really cool innovation there is the ability for delegators to override a validator. So there's a recent SGP 002, double disinflation. That one passed by the thinnest of margins, and the reason it's it passed was that some delegators overrode their validators and said, we disagree with how the validator voted. We prefer a yes or a no. And yeah, it turned out to be great participation. A really, I'd say, successful process.

6:40Unidentified speaker (speaker_3): Awesome. Yeah. And are there any upcoming SGPs that would be worth flagging for our audience?

6:46Unidentified speaker (speaker_0): Yeah. So um one kind of big process point that came out. of this last SGP round was that validators had some confusion about what abstain actually means. So SGP one laid the groundwork for how the process works. One of the pieces in there was that essentially abstain more or less meant that you're voting no. Because the way it works in order for a proposal to pass, yes is the numerator, yes plus no plus abstain is the denominator. Essentially what that means is the yeses have to be 66.6667% of all votes. And validators didn't like that. There's been a lot of feedback there. Uh so there'll be a new SGP coming out that tries to fix that. And we're still debating about what the exact rules should be. We actually have a a governance discussion later today where we're gonna hash some of that out.

7:38Unidentified speaker (speaker_2): Awesome. I love the fact that you called out the nuance that a delegator has the ability to override a validator note. But it's in some ways, drawing a comparison to the TradFi world, it's essentially like when you have proxy voting and an individual shareholder may get the opportunity to vote. They may or may not exercise that, but if otherwise they might delegate that to an institution that might be holding their units on their behalf. So I think that expression is really important and interesting. Are there any other chains you're aware of that actually allow for that type of delegator to override a validator in any governance? I don't know that there are.

8:13Unidentified speaker (speaker_0): Not that I'm aware of, no. Um this was really the biggest point of feedback from the earlier processes that we had. We have a process for iterating the core protocol called SIMD Solana Improvement Documents. And they're we essentially use used to vote on SIMDs. So there was a previous SIMD called 228 that also tried to address inflation. That one lost. And a lot of the feedback was that delegators preferred to override their validators, but they didn't have a method to do it. That's kind of the feedback that came about. And I think that one of the strengths of the Solana community, the Solana validator set, is they want to iterate quickly. So within a year, we had a new process that allowed delegators to override people, override their validators specifically. And I and I think that's a really good foundation for governance to be built on, right? Because validators ultimately cast the vote, but the stakers who hold the soul have sort of the override power, like you said, like a proxy vote.

9:06Unidentified speaker (speaker_2): Very interesting. So you really appreciate going into that in that detail, but maybe we can sort of zoom up a bit and talk about some recent changes that have affected the Solana blockchain. So we had a protocol upgrade called Alping Glow. You know, we're partway through that. Could you talk to us a little bit about what Alping Glow does, you know, where we are phase one versus phase two? And when it's completely implemented, what might be the major changes that users of the blockchain will realize?

9:31Unidentified speaker (speaker_0): Yeah. So Alping Glow is the biggest change to the the Solana protocol in the history of Solana. And you can think of this if you're familiar. with Ethereum, like the merge, right? It's a huge change. We've done this in a little over a year. Really, I think just amazing engineering achievement. What the end user can expect is faster finality, right? So Solana tries to be the fastest, cheapest blockchain around. AlpenGlow is live today, the Alpine Alpen Glow consensus layer. And it allows for about 150 milliseconds in finality. So that's on the level of like web two clients or web two applications, right? Where you can go click a button and then it's done. It's already finalized. And finality, just to like give an idea of what that means, it's the same thing as your transaction being final. So if I want to pay for something at a coffee shop, I can do it on my wallet app, right? And then without having to wait around and check the blockchain a minute or two later, you know. hey, this thing is finalized. I've completed the transaction, I've paid the whatever money, and now I can walk away. Um, which I think is a huge advantage for people who want to use these for consumer level apps.

10:37Unidentified speaker (speaker_2): It's a great call-out. I I think that sort of hits on the medium of exchange use case for cryptocurrencies. And as you're saying it, I was sort of laughing to myself. We've been involved in digital assets of fidelity since 2014. And we started. out with Bitcoin mining as a first thing, and then we gave some people some Bitcoin and some sats to go out and experiment in the ecosystem. What could you do? So you could buy books at a certain university in Boston or Cambridge. And we also had folks who were trying to buy food in our cafeteria using Bitcoin. And a good friend of ours, Alex Thorne, who used to be at Fidelity, is now over at Galaxy. He waited 10 minutes in line to pay for. a banana when he was trying to get it. So it's come quite a long way. And when you talk about the block times, we're now talking about 150 to 200 milliseconds or maybe even faster block times post-Alpen Glow consensus implementation. Is that right?

11:28Unidentified speaker (speaker_0): Uh a little bit. So uh clarification there. The block times on Solana right now are 200 milliseconds, which is by far the fastest of any blockchain. And then the finality is the piece after that block. So 200 milliseconds is the time where validator has to put those transactions in a block. And then once that block is complete after the two hundred milliseconds, all the validators on the network have to vote whether or not they've seen that block. And that's the finality piece. So end to end, let's say if you send your transaction at the very beginning of a block, the total wait time will would be about three 350 hundred and fifty milliseconds, which is, you know, quite fast, but we're trying to push that even lower. We have thoughts and discussions about making the block times even lower, like 100 milliseconds, 50 milliseconds. I've got protocol changes coming up that should help make that possible. We're kind of getting up against the speed of light if we have validators all around the world. So we're thinking about protocol changes where we can group validators in a region so that way we can have Of finality even quicker because there's a lot of validators in the same region to vote at that time. and then we'll move that voting set around the world.

12:29Unidentified speaker (UNKNOWN): Yeah.

12:29Unidentified speaker (speaker_3): And I I think something you you just hinted at there, Tim, was like pre-alpinglow finality was closer to twelve seconds. So that bump to 150 milliseconds and that 350 millisecond, you know, total wait time is a significant increase.

12:46Unidentified speaker (speaker_0): Yeah, absolutely. Um, so there were quite a lot of changes that happened over the past, even just three months that included Alpenglow, which got our finality down to 150 milliseconds. The block times also decreased. So they were at 400 milliseconds and over the first activation, but over the course of this year we've gotten down to 200 milliseconds. So yeah, really iterating on speed and latency there.

13:08Unidentified speaker (speaker_3): Yeah. No, that that's awesome. And something from the node operators perspective we've noticed is pretty pre-App and Glow. I know with the Tower BFT, like validators had to vote on chain, right? Every every slot or every block. Um and that added significant, let's just say overhead to the network and post first phase. A lot of that strain is going away and I I believe's helped to to streamline a lot of the voting that occurs on chain.

13:37Unidentified speaker (speaker_0): Yeah, yeah. So previous. previously and and maybe a criticism of Solana at times is that the votes are transactions on the network. So a lot of the transactions that some people were counting on chain were actually the votes. And now with Alp and Glow, all those votes are not part of the block. We still have voting on the side, not on chain, of course, but those votes, yeah, again don't count as transactions and really reduces the block size. Now it's it's purely for user transactions, and we've got plenty of headroom, right? We've got a hundred million CUs, which gives us a lot of space to have a lot of transactions per second. We've hit 5,000 or more sustained transactions per second over the year, and we have plenty of room to increase it more. In the next year, we're planning to really increase Increased capacity quite a lot.

14:22Unidentified speaker (speaker_3): Great. Yeah. And we know the next phase is called rotor for the phase two of Alpine Glow, and it has to do with improving how blocks get propagated. Could you speak a little if you know the timeline or like what we should look out for over the coming months for rotor?

14:38Unidentified speaker (speaker_0): Yeah. So there's been discussions in the Alpenglow white paper about rotor. The idea is essentially to have more of the bandwidth for those treads, essentially the pieces of the block, to be handled by a single node. And the idea is really to take full advantage of the bandwidth that these validators have in the data centers. There's no specific timeline yet on when rotor is going to roll out. We have some higher priority things that we're really going to focus on at the beginning of next year. I I mentioned it a little bit or alluded to it, which is the GeoWare leader schedule. And then really what we want to focus on quite a bit is multi-concurrent proposers. So there was a talk. at a previous event before Breakpoint called Scale or Die, uh, where Roger Wattenhofer, the researcher at the Solana Foundation now, proposed a protocol change which allows multiple leaders to be creating blocks at the same time, and then some rule for merging those blocks. And we think this will give us some really good advantages in terms of eliminating the ability for a leader to take a minority. monopoly on the block building, right? We see some problems on chain with one leader having full control of the block for some amount of time. That's also one of the drivers why we're trying to get block times lower and lower to limit that monopoly. But we believe multiple concurrent proposers is going to really make certain things that validators are doing you

15:57harder with uh competition.

15:59Unidentified speaker (speaker_2): So when you talk about the multiple concurrent proposers you're talking sub 200 milliseconds or around 200 millisecond time frames for adding transactions. It's a really a lot of sophisticated compute that's going into running this network. What has the foundation been doing to help expand the number of validators that are operating? Or frankly, you and your role, how do you help people understand how they can keep up? Because you know, for a firm like Fidelity, where we're operating in this space, we can be, I'll say, relatively nimble. But I think for a lot of operators, they may questions is the barrier to coming in and running a validator something that they can handle? So anything that you could share on that line would be really interesting.

16:41Unidentified speaker (speaker_0): Yeah, I would say that, you know, maybe there's a bit of misinformation there as well. I think that, you know, most Solana validators operate in a data center typically bare metal compute. And that that can be a hurdle for some institutions, right? I think a lot of people are very used to using AWS for everything. A lot of Ethereum operators, for example, also run on AWS. So I think people are comfortable with that. Uh a lot of like web two practices carry over there pretty easily. But the biggest constraint there is often bandwidth and the cost of bandwidth, to be frank. So I think the biggest constraint on Solana going for forward will be bandwidth. A lot of my role is helping teams get comfortable with transitioning to bare metal operations. But having said that, the core developers at Fire Dancer and at Anza are working all the time to improve the performance of the clients. Both Anza and Firedancer gave a talk a few days ago at Scale or Die, and they were showing off how little memory they use, how little CPU they're using. So the meme of the high compute costs on Solana are really going away because the engineering is just so good. I mean, two years ago the requirements were much higher than they are today. A lot of people on mainnet can get away with 128 gigabytes of memory and probably lower and lower as improvements are made.

17:55Unidentified speaker (speaker_2): That's great to hear. And it is interesting. And it I I think it's been for me as someone who's in the space looking at a number of different technologies and helping to try and coordinate business use cases. You know, one of the things that really appealing about Solana is the is that transaction throughput speed and making sure that we can participate in an ecosystem that's operating at that level and sustaining on our end, because frankly, uh speed is really important, but security is is most important. So I think that's one of the reasons why we're asking this. And you mentioned uh some of the client providers. Um, Kevin, could you talk just a minute about some of the conversations that we've had historically at Fidelity around validator client diversity, what that means to an ecosystem. And then Tim, I'd love to hear your feedback on that as well.

18:39Unidentified speaker (speaker_3): Yeah, definitely. A lot of our research in FCAT actually started with Ethereum. And you know, my area of a focus uh began with running uh Ethereum validators, RD around the Ethereum ecosystem. And yeah, I think Ethereum, a couple, you know, as of or post-merge, you know, had a good healthy client set. You know, and today I think there are four or five different like base clients a validator can choose from. And I think where Solana has made some really good advances over the years is now adding additional choices. for operators in addition to Agave, sort of the base client. So yeah, I think from like a research standpoint, we're constantly scanning what other clients we can test out and almost play with to measure uh different on-chain performance metrics as a validator. There's also the concept of a scheduler, right? So how blocks get ordered by the validator software you're running.

19:39So I think for us going forward, we're looking to test, experiment with different clients and hope that diversity continues to move in the right direction.

19:47Unidentified speaker (speaker_0): Yeah, yeah. Just to add to that, I would say that um client diversity has been a focus of ours for quite a few years now. The Fire Dancer full client is live, available. I don't know the latest numbers because it's been changing day by day, but they've had over 16% of stake for a bit now and are growing. So I I think it's pretty fair to say we have two very strong clients on Solana. And then there's a third. So just at the event yesterday, the Mithril client announced that they are on mainnet and running on a a validator right now. They're still in development, they're still working out some bugs, so it's not really a a full rollout yet, but there's a third client coming on Solana, and we have plans for more. So I I think that from a network securities perspective, this is really helpful because we have multiple clients implementing the same protocol with different supply chains. You know, a large part of my job nowadays, and I think a large part of everyone's job in crypto is security. So we really have to think hard about how we prevent attacks to the network, right? And one ways one of the ways you mitigate that is to have different supply chains, different languages that you use for all these different clients. And hopefully if there is some inconsistency or if there is an attack, all the validators can switch to another client and the chain just keeps going as usual.

21:01Unidentified speaker (speaker_3): The figure we look at is the cost disrupt finality for the networks around that 33% mark or what the community calls a super minority. There's also a 66% mark, right? That is a little more problematic. if there's a larger majority that acquires stake. So we talked about like I suppose clients, but what are other areas you think the foundation is keeping an eye on and where, you know, from a RD perspective, we should be assessing to minimize any single points of failure across the network?

21:33Unidentified speaker (speaker_0): Yeah, so this is something that our team pays quite a lot of attention to, right? I think that a lot of people focus on. validator count. And I would say that's it's obviously an important metric. We don't want to have five validators on the network. But I would say that there are quite a few other metrics that matter quite a lot. So for example, we look quite a lot at the diversity of infrastructure, right? If all the validators are on a single network, like on AWS, if AWS goes down, we've got a lot of problems. So we really look at and at sometimes incentivize even people to switch to different providers to make sure that we don't have a single point of failure there. Another thing that we're really concerned about is geographic diversity, right? We don't want all the validators in one region. Let's say there's some natural disaster there, which has happened in the past. We've had a data center fire in the past that has taken out a whole data center and uh you know cause caused some minor disruptions on the network. but nothing that took the chain down. So lots of things that we're really paying attention to beyond just the the top level metrics that a lot of people look at. Obviously Nakamoto coefficient is still important. We still want the stake distributed across many different uh actors on the network to make sure that no one can have that supermajority. Uh one thing that's actually changed with Alpenglow Alpen Glow is that you mentioned 33%. That's actually um a tower BFT number with

22:52Alpen Glow now. Theoretically it's 40%, right? So it it works on this model where optimistically 80% of the network can vote on something, but there's a fallback down to 60%. So the the chain will progress if only 60% of stake is online. Uh and that 60% is required for finality. So, you know, working to uh basically to um uh minimize the number there a bit to help encourage diversity.

23:18Unidentified speaker (speaker_3): Yeah, that's a great call out. And certainly many layers to decentralization that you mentioned. Yeah.

23:24Unidentified speaker (speaker_2): So, Tim, we talked a little bit at the beginning, the fact that you've been meeting with folks prior to this conference, giving some talks, interacting with in different forums. What have you seen so far to date that you're most excited about as we're heading into the first full day of breakpoint?

23:38Unidentified speaker (UNKNOWN): Yeah.

23:39Unidentified speaker (speaker_0): So um some background there. I I would say four or so years ago, we had this idea to have a a validator meetup. And at the time I thought people were thinking along the lines of like hang out at a bar, say hello, get to know people one on one. But the validators really took it and ran with it and created like a whole separate conference track that they put on themselves and and work with the Solana Slana Foundation to make into a I would say a first-class production. So that I believe was four years ago. The first it's called Block Zero Conference. There's a sort of sister conference that came out of that as well. The name has changed a little bit over the years, but now it's called 200 milliseconds. Essentially, trading is the overall theme there. And then about a year or two ago, I'm a little fuzzy on dates, but we introduced a conference called Scale. or Die. And essentially the idea there is a really engineering focused conference that's all about what we're shipping in the next year. So all of those conferences happened before breakpoint and so much good energy. It's Scale or Die, in fact, was oversubscribed. People were waiting outside for an hour, unfortunately, but there was just so much interest that it was hard to get all the people in. We were over capacity for the venue. So anyway, lots of great things going on. I I would say the things I'm most excited about are I have a bias because I'm much more at the protocol infrastructure layer, but I

24:58think the the protocol changes that we've got coming up. I think, yeah, MCP is really exciting for me. I think both client teams on the network right now, Anza uh Or Agave, I should say, and Fire Dancer, have worked to make the clients so, so fast and so efficient on research. resources that we have so much room to grow. And in the next, you know, in the next year, I think that we're really gonna surprise a lot of people with how fast and how efficient the clients are and and how much room we have for block capacity. Because, you know, one thing we talk about quite a lot is that we're not, we're not building for the average case on chain. We're building to double or triple the most traffic we've ever seen. And yeah, that's what we're hoping to have the capacity for in the next year for traders and for asset managers and just institutions on chain in general.

25:44Unidentified speaker (speaker_2): It's really exciting to hear. And I'm, you know, my background was in the traditional financial services, managing everything post trade around the globe. And I what I'm seeing a lot of is this push towards asset tokenization, whether it's native or some representation of an off-chain off chain asset. And I think for some of these participants who want to see that at scale, they're looking for speed. And I think some of the advances that the Slana Foundation that you all are doing will certainly help to advance that understanding or maybe even lower the barrier of disbelief that exists out there. I I can recall a long time ago talking to someone saying, I think we're going to be seeing assets trading on chain like stocks and bonds. And the person just shaking their head and now we see the world's largest security securities depositories working with others to try and make that happen. So it's, you know, as the saying goes, gradually and then suddenly. I think we're getting closer to the suddenly, but it's really exciting to hear about what you all are working on, how that's benefiting the ecosystem, security, accessibility, and quite frankly, the speed of which transactions can process now is is mind-blowing. mind blowing. So I'm really excited about it. Kevin, because before we wrap up, do you have any other questions or anything that you'd like to share from your research or what you're most excited about?

26:57Unidentified speaker (speaker_3): I'm generally just, you know, excited for all the upgrades you mentioned. You know, we'll we'll definitely test out some of the clients you mentioned earlier. It's my first breakpoint. So I'm just excited to engage with the community, get out there, listen to some of the talks. But really appreciate your time today. Yeah.

27:14Unidentified speaker (speaker_2): it's it's my first break point as well. So, and what a way to kick it off. Uh sitting down, speaking with you about what you've been doing and how you're helping this ecosystem grow. That's it's really inspiring. So really appreciate your time. And when it comes down to Fs at FCAT, we're gonna continue to be monitoring this ecosystem, bringing back a lot of education from the experience we're going to have here at Bray Point, and really looking forward to summarizing it in the next crypto brief. Thank

27:38Unidentified speaker (UNKNOWN): you.

27:43Unidentified speaker (speaker_0): Crypto as an asset class is highly volatile, can become illiquid at any time, and is for investors with a high risk tolerance. Crypto may also be more susceptible to market manipulation than securities. Crypto is not insured by the Federal Deposit Insurance Corporation or the Securities Investor Protection Corporation. Investors in crypto do not benefit from. the same regulatory protections applicable to registered securities. Past performance is no guarantee of future results. This podcast was produced by the Fidelity Center for Applied Technology, also known as FCAT. FCAT does not offer digital assets nor provide clearing or custody of such assets. It is for informational purposes only and is not intended to provide tax, legal, insurance, or investment advice. and should not be construed as an offer to sell, a solicitation of an offer to buy, or a recommendation for any security or other asset by any fidelity entity or third party. Views expressed are as of the date indicated, based on the information available at the time, and may change based on market or other conditions. Unless otherwise noted, the opinions provided are those of the authors and not necessarily those of Fidelity Investments or its affiliates. Fidelity does not assume any duty to update any of the information. Fidelity and any other third parties mentioned in the podcast are independent entities and are not affiliated. Mentioning them does not suggest a

29:02recommendation or endorsement by Fidelity. This information is not intended for distribution to or use by any person or entity in any jurisdiction or country where such distribution would or use would be contrary. to local law or regulation. Persons accessing this information are required to inform themselves about and observe such restrictions.

29:22Unidentified speaker (UNKNOWN): Third

29:23Unidentified speaker (speaker_0): party trademarks appearing herein are the property of their respective owners. All others are the property of

29:28Unidentified speaker (speaker_1): FMR LLC. Copyright 2026, FMR LLC, all rights reserved. 1040156.

29:33Unidentified speaker (UNKNOWN): one

29:33Unidentified speaker (speaker_1): zero four zero one five six.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights