Austin：让加密更快、DoubleZero 代币、Solana 交易及更多｜TG Podcast

Austin Federa, welcome, man. It’s a pleasure. How are you?

Good, man. Thanks for having me on.

Of course, dude. I appreciate you coming on. I can imagine it’s a busy time for you right now, huh?

It’s a busy time, but honestly, when is it not?

Yeah. When is it not, dude? Welcome. I’m honestly excited for this one. I think for a lot of the Solana on-chain stuff that we do, this is a fun switch-up, and I think it’s just perfect. It just works.

Do you want to start with a quick introduction to who you are, and then we can get into some of the DoubleZero stuff?

Yeah, sure. I’m Austin Federa. I’m one of the 3 co-founders of DoubleZero. I got into crypto in late 2017, working for a company called Republic, which was launching Republic Crypto back then. They did a bunch of fun ICO stuff back in the day.

Then I worked for a few infrastructure companies, notably Bison Trails, which got acquired by Coinbase and became Coinbase Cloud. After that, I spent the last 4 years at the Solana Foundation as the head of strategy there.

You were a really early Solana employee.

Yeah. There were definitely earlier folks than me, but I joined about 9 months after mainnet and about 2 months after they laid off half the employees.

Whoa. So it was not a consensus move back then. When did you leave Solana?

I left in November 2024, so November of last year.

You left at a crazy time, dude.

Yeah, but at the same time, you’ve got to figure out when you’re going to move on from a place, right? Solana is incredible. The Solana Foundation is an incredible place, but 4 years is a long time. You get the itch to be a founder. You get an idea that you can’t get out of your head, and you’re like, “Solana’s doing great. We got through FTX. We got through the bear market. They’re fine. They’re going to do great. Don’t have a big enough ego to think you’ve got to stick around and help these guys.” Solana is doing incredible.

This is a little off-topic lore stuff, but you were there during FTX?

Yes.

Did you think it was over?

You know, this is kind of the funny thing: I don’t think any of us thought it was over.

That’s awesome.

I think it was one of those moments where there was this sort of collective reaction. It wasn’t even around the room. This happened, if you guys remember, right at the end of Breakpoint. We came off this awesome high of our second conference, with the most people we’d ever had in Lisbon. I had 4 days of vacation lined up, and suddenly it was like, “Oh, the world of crypto just completely collapsed.” I didn’t leave my hotel room for 3 days, pretty much.

I don’t think anyone thought it was over because, fundamentally, Solana was super strong. The tech was super strong, and the community was strong. There wasn’t really anywhere else people could go to build anything remotely similar to it. So it was just, “Hunker down, treat people like people”—especially the users, builders, and holders who were going through their own thing—and we’ll get through this.

In the moments after FTX, was there a Raj-and-Toly all-hands war-room call?

Oh, yeah.

Yeah.

It was insane there. It was genuinely one of those things where, for probably the 8 weeks following, I don’t think any of us slept more than 3 or 4 hours a night.

I wish I could have been a fly on the wall for some of those meetings. That’s awesome.

It’s kind of like when you talk to folks who’ve been through real hard times. They’re like, “I wish I didn’t go through that, but I’m also really glad I went through that.” That’s very much what it was like for all of us coming out of FTX, especially a year later at Breakpoint the next year, being like, “Oh, wow, 4,000 people showed up again for 4 days of programming.” You started to see the market start to recover. But yeah, it was a rough year.

Yeah, that’s exactly how I felt, except it was holding Truth Terminal Goat down 90%. But same feeling.

So you left in November 2024, which was honestly a crazy time. It was a crazy time for me, too—that’s when I started streaming. It was an awesome time, AI season on Solana, and you had the founder itch that you couldn’t get out of your head. Then you went and started DoubleZero.

I’m going to be honest: I don’t really know too much about what DoubleZero is. That’s why we have you on stream to explain it to us, which is awesome. That’s why my job is incredible. Can you give us the breakdown? How did we get here, and what are you working on with DoubleZero?

Yeah. DoubleZero is a totally new type of blockchain project. We’re not a new layer 1, and we’re not a layer 2. The whole purpose of DoubleZero is to build a new internet optimized for high-performance systems like blockchain.

We’ve never had something like this for crypto before. The traditional tech world and traditional finance have been using private, dedicated networks that exceed the speed of the public internet for decades at this point. Google, Amazon, Microsoft, Facebook, and OpenAI are building their own fiber networks.

This is also true of DRW, Cumberland, the firms depicted in *Flash Boys*, and all of the high-frequency trading folks. They also run their own private networks because they’re faster than what the internet can provide.

Blockchains have always just run on the public internet because it’s expensive to build one of these networks. Quite frankly, for a while, blockchains weren’t fast enough to use them properly. But it’s 2025, and we’re in a very different place. We’ve got Firedancer, we’ve got Anza, and we’ve got Aptos, Sui, and a bunch of these L2s. They’re running into the limits of what the public internet can support. So we figured, hey, it’s time to go try and build a new internet.

So who’s running these private networks, and for what purpose?

We have 12 independent contributors that are all contributing fiber capacity to the network. It’s all sorts of companies. There are some trading firms, both in traditional finance and new finance, so we’ve got Jump, DRW, Cumberland, and Galaxy.

Then there are actual people who own data centers and whose businesses are selling servers, like TeraSwitch, Latitude, and Servers.com. We’ve also got some VCs involved. The way they might say, “I’m going to go build a Bitcoin miner” or “I’m going to run a validator for a network,” they’re actually going out and getting fiber to contribute to the DoubleZero network.

You mentioned OpenAI and Amazon. There are a bunch of existing companies that are basically running these networks that aren’t the public internet. For what purpose? Why, if I were OpenAI, would I want some ridiculously high-speed network?

The answer is crazy, because I’m going to say it and you’re just going to be like, “That can’t be the reason why.” It’s to sell ads.

Okay. How so?

Facebook runs a giant, multi-billion-dollar private internet to sell ads. What they need is for anywhere in the world, when you pull out your phone and start scrolling Instagram Reels, everything loads instantly and you have a great experience. There’s no buffering or anything like that.

The more you scroll, the more ad impressions and ad units they can get. What they realized was, especially when they were going into new markets, they could basically wire up most of the Philippines with fiber so they could sell ads to people there.

Really?

At the end of the day, they’re trying to expand the market they can reach, which is crazy when you think about it. That’s actually the thing.

Google built its network for a slightly different reason. Google wants a 50-millisecond response time anywhere in the world, so they have to have all their different data centers linked by private fiber. That lets them update search results and other things in multiple places.

But at the end of the day, all of this is because the internet is too slow for how big these businesses have gotten.

Can you, as a regular consumer, get access to these networks? I used to bot sneakers, and we would run intensive AWS servers and run proxies on them to get better speeds and whatnot. Is that the same thing, or am I completely off?

No. What you were doing is a few levels above where you start optimizing what people are looking at. If your entire business is this, you’re going to figure out where in the world the web server hosting the sneaker drop is physically located. Is it in Tokyo? Is it in Chicago? What Amazon region is it in?

Then you’re going to try to brute-force a bunch of different AWS instances. You’re going to launch a ton of them because they don’t tell you what physical server you’re getting. You’re going to keep pinging this web server and try to get the lowest latency you possibly can. Then, the minute the drop goes live, you’re just going to blast it with all of your sales requests.

This same thing is what high-frequency trading is. The difference is that what they’re doing is arbitraging between 2 markets. You’re trading on the New York Stock Exchange and a price moves, right? Something happens—there’s a news event—and you’ve got to reflect that change somewhere else where you’re trading.

Maybe Chicago, the CME, maybe it's Tokyo, and you're trading on the Nikkei. The faster you can get that data from point A to point B, especially if you can beat your competition, you're going to be able to trade before they know it.

It's sort of like if you could go through an entire day of trading on crypto, then say, “You know what? Let me rewind 24 hours and trade again.” Knowing exactly what's going to happen in a 24-hour market, you'd beat the market dramatically. You'd have all this information that no one else has. Being faster in trading is basically as close as you can get to having a time machine.

You kind of just touched on it. Could you break down a little bit how high-frequency trading works and why certain firms are better than others? What do they need to be successful?

Yeah, there are 2 main components to making high-frequency trading work. There's your trade intelligence, your sort of signaling—the idea of how well your quantitative trading model can predict what's about to happen in the market. The second one is how quickly you can respond and do something about that.

These 2 are related, but it's like having a supercomputer cluster that's pulling in data from all sorts of different sources all over the world. It has price feeds from stock exchanges, but it also has news reports and all these other types of things. When it sees something that it thinks might move a price—maybe the factory that makes 30% of the world's iPhones just caught on fire—that's going to have some pretty serious impacts on a whole bunch of tech stocks.

Or there's a rumor that just hit the wire about a new tariff going in. All of these things can create market movement. Then the question is, how quickly can you actually act on that? Knowing something is great, but if someone else is going to beat me to actually being able to trade on it, that knowledge isn't actually useful yet.

The second component is all of these global fiber networks and microwave towers, and even shortwave radios, that they use to trade nowadays. That's the world of high-frequency trading. It's great to be smarter than your trading counterpart, but it's better to be faster. If you can only pick one, you should pick being fast, because it's very hard to lose money if you're faster than your competition.

You can be smarter and still make mistakes. That happens all the time.

I just watched Margin Call this weekend. It's like, be smarter, cheat, or be first.

I'd like to think there are some smart people at this firm. I would rather just be first.

I love that. That's awesome. My obvious question for you is this: I trade on Solana all the time. A lot of people in this chat trade on Solana all the time, and they're probably listening to this thinking, “Okay, cool. New, faster internet sounds awesome, but Solana's really fast. I get transactions through on Solana in a couple seconds. It's instant. I've never had a problem with it. Why does anyone actually need this?”

Solana is super, super slow compared to TradFi. 400-millisecond block times are unfathomably slow. Now, it's super fast for crypto, but it's pretty slow for TradFi.

That's kind of where the delta comes from. It's the same way that if you're trading on a market on a blockchain with 30-second block times and you're trading on Solana, Solana's going to run laps around you. You're always going to have better pricing because market makers can get their quotes in more quickly. They can cancel their quotes more quickly. The whole block just moves faster, so the market moves faster.

With high-frequency trading, people think it's sort of like, “Oh, it's manipulating the market or something.” What it's really doing is bringing the market back to the true price as quickly as possible. You can think about it as having a blindfold over your eyes while you're trying to walk a tightrope, and you get a little bit off. It's about how quickly the market can push you back to whatever that true price is.

This is where high-frequency trading is actually a plus for markets. Why do you get such good prices on Robinhood? Why do you get such good prices nowadays when you're trading? It's because there's so much high-frequency trading going on in the background. I don't really care if I lose out on a tenth of a penny on this trade. I care that I don't have to spend a dollar to do the trade. That's kind of where all that comes from.

Now, on Solana, Solana's fast, but there are a whole bunch of other things that go into whether you can get your transaction in. For example, I'm sitting in New York today. If I want to trade on Solana, what happens? I open up Phantom or Solflare, which is talking to an RPC server somewhere in the world. First, we have 1 latency hop there.

I hit trade. I have to get a price from the blockchain. I hit trade, it sends it to an RPC, and that RPC then has to send it somewhere in the world to wherever that block is being built. Maybe that block is being built right next to my RPC server. Great, I get an almost-instant response.

Maybe that block is being built in Singapore, and now I have to send my data all the way from New York to Singapore. On the public internet, that can take 300 milliseconds or 2 seconds. It entirely depends on the state of the public internet at the moment. There's a huge amount of variability in how quickly your trade actually goes in.

When you're saying, “Hey, it takes a few seconds,” that's awesome. From a retail, direct-experience standpoint, that is awesome. For a market maker, they could have just lost hundreds of thousands of dollars because the price on Binance moved, and suddenly they're selling something that's worth $1 for $0.90.

That's where, when we can bring faster, more high-performance networking into this and accelerate networks like Solana, you're going to see better pricing as a user. You're going to see less than 10 basis points of slippage.

You've seen this already start to happen with some of the prop firms.

Yeah, 10 basis points is a big deal. It's not nothing at all. I think we can get pricing on-chain on parity with something like the Binance order book.

On a CEX? That's crazy.

Yeah, which is awesome. Then we can bring price discovery back on-chain, and that's what we're all in here for.

My understanding is that you're basically selling this super-fast, low-latency internet service to blockchains, and you could substantially increase the speeds at which blocks process and transactions go through. Who are your customers? Who's actually using this?

Yeah, there are a few different groups. The first are the validators themselves. They're one of the most important parts. If they're operating on this and using the DoubleZero network, they're going to be able to process transactions faster, build bigger blocks, and build better blocks.

Then the developers at Solana are going to say, “Look, we have all this extra capacity. We can make Solana even faster than we could otherwise.” That's kind of that flywheel effect: we make the network stronger, and then we're able to make Solana faster because of it.

The second component is the traders. These are market-making firms, professional traders, and just people who are spending a ton of time actually trading and in the trenches. They're looking at ways they can optimize their latency.

Especially if they're doing arbitrage between a CEX and a DEX, that's where DoubleZero is the fastest path to get from the Coinbase order book, the Bybit order book, and the Binance order book onto Jupiter, onto something on-chain. That's where we see a ton of value over the long term: we can bring all the positive dynamics of high-frequency trading, price discovery, and arbitrage to on-chain markets as well.

This is all built in advance of multiple concurrent leaders on Solana, because that's when this stuff gets very cool, in my view. We can build way better markets than we can build today, and it allows DeFi developers to take advantage of stuff they really didn't think was possible.

Does your customer just have to be a Solana chain-wide integration, or can it be firm by firm? Can a trading firm tap into your network and just get better speeds than everybody else on Solana?

Yeah. The network gets more useful the higher percentage of stake is connected to it. Ideally, we want to get 100% of the validators someday connected and reachable through the DoubleZero network, but an individual validator will benefit from being on DoubleZero, and an individual trading firm will also benefit.

They may not benefit 100% of the time. If you are, for example, a market maker and you have to get a cancel order to a leader, is that leader on the public internet? DoubleZero isn't going to help you then, so you route through the public internet.

But if that leader, that block builder, is on DoubleZero, you suddenly can reach them much more quickly than you thought was possible before. These folks are very sophisticated, so we may actually see different price quotes and different spreads when different validators are building blocks, based on whether they're on DoubleZero or not.

Now, someone’s got to build all this out, but there’s so much revenue and economic activity on a network like Solana that everyone is optimizing as much as they possibly can. The folks at Temporal, with their PropAMMify, have done stuff to on-chain pricing that people didn’t think was possible until cancel prioritization came in.

We’re just seeing this complete renaissance of high-performance DeFi, and our whole thesis is that we’re building this all on a broken foundation. The public internet is not the right way to build this stuff anymore.

Okay. So DoubleZero works better the higher percentage of validators are tapping into the network. How do you make money from that?

Validators will pay basically a seat fee to operate on the network. But we’re a token project, right? The token is the ultimate thing of value here. This is the design: validators pay in 2Z tokens to use the network. About 50% of those fees are burned, and the other 50% are distributed to the people who are actually providing fiber.

We at the DoubleZero Foundation are totally out of the loop, the same way the Solana Foundation is totally out of the loop on any transaction. We’re a nonprofit Cayman foundation. We’re doing the classic blockchain foundation thing.

So, hypothetically, if every validator on Solana was using DoubleZero, what percentage of the token supply would be burned?

We’ve got to say, okay, so if we’re paying, we have 100% of it. The fee structure is about 5% of block rewards. Block rewards are pretty variable; it really depends on the epoch. In general, we’re talking probably about 250 SOL per epoch worth of DoubleZero tokens, right? It depends on what the exchange rate is and all these sorts of things.

It’s a little bit of a complicated calculation, but this is the classic model: with enough demand through the system, you start to see decreasing supply. We do not have inflation on the protocol today. There may be some staking in the future, but it’ll be a very low staking reward rate, if so, in part because we have this whole different reward model called proof of utility, which is different from what you see in other blockchain systems.

Got it. So you guys can’t really accrue revenue from the validators. Your main model is that you have to sell tokens to make money from this.

Yeah. We’re a blockchain foundation, right? The Solana Foundation has no other revenue streams either, really. That’s kind of the way it works.

For the contributors, the folks who actually have the fiber, the fiber is really expensive. We’re talking tens of millions of dollars a year for the cost to actually maintain this network. They’re the ones who actually need the revenue coming out of the network. That’s where 50% of the tokens get burned, and the other 50% are distributed to the folks actually providing fiber capacity.

Got it. How much does Solana make on network fees?

Solana Labs and the Solana Foundation make $0 on them.

Oh, damn.

Yeah. The reason no one owns equity in a blockchain foundation is that it’s just an entity that loses on it.

Eventually, if the token price stops going up, these things go out of business because they’ve sold off all of their tokens. But this is by design. You don’t want a giant for-profit entity that’s making a lot of money at the center of a blockchain network. You want it owned by the people who use it. You want it owned by the validators and the contributors. You don’t want it owned by a foundation.

Got it. Okay, makes sense. Your TGE is tomorrow. Is that correct?

Yeah, tomorrow.

That’s exciting.

Super exciting. 9:00 a.m. Eastern.

What do we need to know? What’s the plan? How are the tokenomics broken down? Incentive rewards, airdrops—what does the setup look like?

We have about 20% of the Solana validators by stake running on DoubleZero today. The mainnet beta—the network—is actually stood up and launched at the moment. We’ve got a bunch of validators operating on it.

The thing that happens tomorrow is fees start. Those validators have to start paying in, and that’s where the DoubleZero token going live matters, because the token is required to use that network. It’s a key part of the economic design.

The token will list on exchanges. Exchanges do their thing. Trading does its thing. We kind of go from there.

It’s a little controversial, but we’ve been clear about this from the beginning: we don’t have any big airdrop or anything like that as part of this. We’re really focused on proving out the value of the network to start.

I think it can be really effective if you have a project like Pump.fun, something that’s got a huge amount of retail activity and those sorts of things. Airdrops can be awesome. They can be a really effective tool.

We’re more of a back-end protocol. I think it’s exciting to people because there’s a lot of potential to make blockchains much faster thanks to DoubleZero. But it’s kind of like an oracle airdrop or something like that, you know? The core users of the thing aren’t necessarily the airdrop market, at least to start.

Okay, that makes perfect sense. If you’re a buyer and holder of DoubleZero on the open market, can you explain what the bet or thesis is that you foresee coming true?

Sure. We’re going to need more transactions in the future than we do today, and we’re going to need faster execution environments than we have today.

You can think about it this way: to date, all blockchain performance is sort of limited by the public internet. If you think we’re good and we have as many transactions nowadays as we need, and Solana is just as fast as it ever needs to be in the future, cool. That’s a thesis you might have.

I personally don’t share that thesis. I think we’re in an arms race for speed and performance, and the best way to address that is to go to the lowest layer of the system and say, “Hey, look, we’re going to build a parallel internet that has technologies that aren’t possible on the public internet.”

The token has some properties baked into it where you can have decreasing supply over time based on the token burn on this system. There are a lot of areas here that are very hard to talk about on a livestream because they run into legal landmines pretty quickly.

One of the awesome things that happened this week is that we actually got a no-action letter from the SEC.

I was going to ask you about this. What does that even mean?

This is a very niche legal determination. Basically, we went to the SEC and said, “Hey, we want to do this stuff. Do you see any problems if we do this stuff?” This is called no-action relief.

We went to them and said, “Hey, look, we have this token we’re going to put on the DoubleZero network. Here’s how the network works. Here’s the utility behind this token. Do you think this is a security? Do you think anything we’re doing is going to run afoul of securities law?”

Specifically, this was around distributing tokens to contributors as rewards, paying for the fiber in tokens, and the staking component as well, which will go live eventually in the future.

The no-action letter is the first time the SEC has issued one in a very long time for a crypto project. What it basically said is, “We think what you’re doing is fundamentally not regulated as a security.”

That’s a huge thing. It’s a very niche legal thing, but this opens the floodgates to a whole bunch of different types of blockchain projects looking to build and launch compliant tokens in the United States and operate in U.S. markets.

I’m really excited about this not just for DoubleZero, but because I think this is a pretty big unlock for an industry that has existed in a gray zone for a while. That gray zone just got a little bit smaller. It’s a little bit more clear what you’re allowed to do.

Congratulations. That’s awesome.

Thank you.

That’s a ton of work from our general counsel. She’s been working on this thing for months.

She probably got a raise after that one, I’m sure.

Give her any ideas.

Yeah. So you guys already have 20% of Solana validators in the network using DoubleZero. That feels hot. How clear is the path to 100%?

Yeah. If you go to 00.xyz, you can basically see a map of all of our fiber in real time and how it’s operating around the world.

This is all contributed from a bunch of independent contributors. You can click on each link, see what the capacity is, and see how much it’s actually being used at the moment. You can dig in and see the validators that are on it.

Our core objective and our core metric that we live and die by, apart from the performance of the network, is whether we’re able to make validators more money. If we’re not making validators more profitable, they may want to operate on DoubleZero because they’d like to support the vision of blockchain at NASDAQ speed, but everyone, at the end of the day, is self-interested.

Not everyone. Many people are interested.

Fair.

And so, yeah, our core value proposition, our metric, is: can we make it more profitable, the same way Jito made it more profitable to run Jito than to run the stock client? We’re doing a bunch of work with Jito and some other folks to make use of this additional capacity and reduce latency to actually increase both the value of blocks.

Just a few days ago, the Firedancer team proposed a really awesome SIMD—I think it’s 293, 292, something like that. What it basically does is remove compute limits from blocks. It basically says, “Hey, validator, build the biggest, baddest block with as much compute as needed as you can.” And that’s really awesome because that’s something the DoubleZero network can really help take advantage of.

So, are validators making more money with DoubleZero? Is it binary? Yes or no?

Yeah, they are. We had a testnet up and running for about 6 months, which is a much smaller network than this. There should actually be somewhere in here where you can press a testnet button. But the testnet was much smaller.

Even on that smaller testnet, we showed that we were reducing skip rates, which is basically when validators had the right to build a block but, for whatever reason, missed building one. Maybe they didn’t have the data in time or something along those lines, and so they lost out on money. We’re preventing validators from losing money, which is basically the same thing as making them more money.

We also see that the median reward rate is higher. Now, these are not 20% increases. There’s a whole bunch of stuff to work on, and as we get mainnet rolled out, more connections and more endpoints will mean that there’s more value the network can add. But today, we are making folks slightly more money, and we hope that is more than slightly soon.

Cool. And so, I guess one of my last questions for you is: obviously, you’re sort of Solana-adjacent, and we talked a lot about Solana, but what about everything else? There’s a Monad in the background, there’s MegaETH, there’s Ethereum mainnet. Can you improve speed on Ethereum mainnet? What about everything else?

Yeah, I mean, look, we are base-layer-neutral infrastructure. We will work with anyone. That is the goal here: to accelerate all blockchain networks and all validator networks. There are L2s we’re already testing with, and there are a bunch of other L1s we’re in conversations with.

Ethereum mainnet and Bitcoin are not really built for performance, so they probably wouldn’t benefit too much from DoubleZero. But flash boys probably would. There are a lot of areas where we can plug in here, even on centralized L2s and something like Hyperliquid. Hyperliquid has a massive state-propagation problem. It’s very hard to run a read node and get data off of it, even though the core validator network itself is doing quite well.

There are a lot of technologies on DoubleZero, like multicast. Multicast is super cool. If I could just delve into that for a second, please. It’s basically this technology where the switches can replicate packets themselves.

Okay.

So what happens today is, I’ve got to send out 2,000 copies of all this data to all the validators in Solana. With multicast, you only have to send that data out once, and then the switches replicate it and bring it everywhere it needs to go.

At the end of the day, it’s like the commercial-industrial version of AirPlay, where you’ve got one stream coming off your phone and it’s going to 10 speakers in your house.

Whoa, that’s pretty awesome.

It’s super cool.

Yeah, thank you.

You can’t run multicast on the public internet.

Word. Thanks for breaking that down. And just so I fully understand: TGE is tomorrow. Basically, the reason the token exists is that validators need to buy and pay with the token to tap into the network, and they want to tap into the network because they make more money. You’re basically giving out half of the rewards back in the token, and then the rest is simply burned in a deflationary mechanism. Is that right?

Yeah, exactly. The whole thing, too, is that we have contributors to this that are like Galaxy. Why does Galaxy care about fast fiber? They’ve got a market-making and trading operation. They also just own a bunch of SOL, and they think that if Solana is 10 times as performant in 2 years as it is today, it’s going to be good for their underlying holdings.

Even if you’re not a validator, even if you’re not someone who runs blockchain infrastructure, if you care about fast networks and think that’s a competitive advantage for a network like Solana relative to others, I think you should care at least a little bit about what DoubleZero is doing.

What’s your comp? What other company does something similar? Maybe it could not be blockchain-agnostic; it could be in the real world. Who’s your closest company?

Cloudflare, maybe.

Cloudflare.

Yeah, something like that—Cloudflare and Akamai. Most of these networks that do exist are very secretive because they’re run by private trading firms. They don’t want anyone to know about what they’re doing.

This is the first time anyone has built a private network like this that has this sort of win-win consortium model, where multiple people are invited to operate on it and it makes everyone who operates on it experience higher quality. Usually, private networks are anti-network effects.

Word.

Which is ironic, but you know, it’s how things go.

Cloudflare is a $76 billion company, according to Google. Why has no one else built this?

It’s very hard.

There are 2 reasons. The first is that, until very recently, blockchains haven’t been fast enough to need this. The second reason is that it’s very hard to build.

We have engineers who specialize in high-performance networks, and they come from everything from traditional exchanges like the New York Stock Exchange and NASDAQ-type exchanges, to folks who have worked for high-frequency trading firms, to people who worked at Amazon building its high-performance network, to folks who come from space companies that have been working on very complicated problems of communicating with stuff in orbit.

It’s really just a quite hard technical problem. The other thing, too, is that modern computer science and software have forgotten how to do this stuff. I don’t mean that as a criticism, but most things in Web2 exist on a very high level.

This is the lowest level you could think of. This is literally cables in the ground. It’s the lowest-level, most boring stuff in a way that you could think of.

So, do you think there’s going to be—maybe I’m off, but at least the use case that we’ve articulated on this stream is that high-frequency trading and trading firms are going to be the biggest beneficiaries of this?

I think maybe it depends. Yeah.

The question I was going to ask was, do you think there are going to be a bunch of these firms that can’t interface with crypto that are going to start with the integration of DoubleZero?

Yeah, I mean, I think so, because right now learning to trade in crypto is a process. When you first started trading, you were like, “Oh my God, what the hell is going on here?” It’s totally different from the equities market.

They have the same version of that, in part because the markets move in such different ways, but in part because, in an equities market, when they say “trade,” the trade just happens. It’s confirmed. Here, it’s like, “Oh man, did the market move between when I hit trade and when it got there?”

There’s a lot of execution risk. Market makers are crazy. Market makers don’t exist in traditional exchanges. They have to exist in crypto because crypto doesn’t have the determinism that you’d expect it to have if you’re coming from a traditional market.

All this stuff comes together to basically make this more attractive. Any network that’s on DoubleZero is going to be more attractive for traditional finance to come build on.

Got it. Okay, this is my last question. This is actually really interesting. I’m actually engaged.

Awesome.

You do a good job of breaking down complex things pretty simply, by the way. Well done.

Thank you.

You have a professor in your future. I can see it.

Don’t tell my dad. He’s a professor.

Yeah, there you go. Okay, you could go technical on this, but can you actually explain how it works? What is literally happening to allow this? Just go one technical level deeper, just for fun.

Explain what literally makes this possible? Okay, let’s start from the network, then I’ll go to how data actually flows through it.

There are cables under the ocean, over mountains, and buried under our feet that are fiber-optic cables. What’s really cool about the physical layer of fiber-optic cables is that a cable laid 20 years ago is still totally useful today, whereas a computer from 20 years ago is probably not useful for anything.

What this means is that folks can go out and get access to parts of that cable, and they basically dedicate bandwidth on it. The DoubleZero network is a network of physically interconnected, long strands of glass that have lasers on both sides blasting light back and forth.

The really cool piece about this is that it’s all physically interconnected. It literally is like there’s a guy plugging one cable into another cable in a data center somewhere, and suddenly we’ve added a new link of fiber to the DZero network.

All of this happens on the public internet too, but the public internet is completely full. There’s so much data moving across it that you get a ton of latency. When we’re talking about this stuff, what we’re really talking about is path dependency: data will always move over the same path, and you always know how long it takes.

Now, a validator connecting into this—when they transmit data over the public internet, if I’m trying to talk from New York to Tokyo, my data might go to Boston, to Chicago, to Atlanta. It’ll bounce all around the United States before it finally makes it across the ocean.

Okay?

And that’s not because the public internet is fundamentally bad at this. It’s because it’s based on how to get the cost of the data as low as possible before sending it.

With DZero, you will have functionally zero jitter. The time when you send a transaction will always be the same if it’s going to the same place. The validators connect through a software tunnel. There’s basically a system on their validator that gives them a connection into DZero and a connection out to the public internet, and they automatically determine which is the best path to send a piece of data.

Wow.

At a high level, that’s how the thing works.

Yeah, that was awesome. And you code—are you an engineer? You coded this? Obviously not.

I’m not. We have an incredible engineering team, so I’m not an engineer on the project.

But you are sort of an architect in understanding how it works.

Yeah. I would describe myself as systems-level technical. You don’t want me actually building and coding anything, but I understand the high level of how it works.

Exactly.

Yeah. I mean, Andrew McConnell, my co-founder and CTO, is the true architect behind this thing. He spent years working at traditional exchanges, and then years working on and running basically high-frequency trading networks as well.

From the engineering side, they’re all super on it. The funny thing is, most of our engineers have no crypto experience.

Whoa.

It’s been a real combining of two worlds: people like me, who have been working in crypto since 2017 or 2018, and engineers who are like, “What is this blockchain thing?” Culturally, I’ve literally gotten DMs before from people asking, “What’s up with these giant penguins?”

That’s awesome. This is sick, dude. I’m not going to lie, I wasn’t sure what to expect. This was a movie. Is there anything you want to sign off with? TGE is tomorrow. You’re on a little bit of a media tour right now, which is awesome. Is there anything you want to leave us with? Anything you want to drop, a final note, whatever?

Yeah. Hey, look. If you’ve been inspired by this conversation and you stake to a validator, talk to that validator. Tell them to look at running on D0ero. And I think we can get Solana up to 1 million transactions per second in the next few years.

Whoa, that’s crazy. A million? What does it do right now?

About 10,000.

Wow. Post-Firedancer, right? It’s 10,000.

Yeah. Yeah. Firedancer can do—you can benchmark Firedancer. It has done hundreds of thousands, almost a million transactions per second before.

Whoa.

But it did it over private networks. It didn’t do it over the public internet.

What was Solana doing in 2023?

Probably about 4,000.

Wow. That’s awesome.

There’s an engineering team working on Solana, both at Anza and Firedancer, and they’re super cracked. They’re really excellent at optimizing as much as they can out of software.

Damn. My goal is to basically put the work back in their camp and be like, “Hey, look, the internet’s the problem now. We’re going to remove the internet from the problem, and then you guys are going to have a whole new set of problems to fix.”

Sick, dude. This was awesome, man. I appreciate your time. Thank you for coming on. It was honestly sick. You do a great job. Thank you.

Appreciate it.

I’m rooting for you, man. See you. Have a good one.

Appreciate it. Peace.