Austin: Making Crypto Faster, DoubleZero Token, Solana Trading and More | TG Podcast
- DoubleZero is building a contributor-supplied private fiber network to remove the public internet as crypto’s performance ceiling. Twelve independent contributors—including Jump, DRW, Cumberland, Galaxy and data-center operators—supply capacity; validators and traders connect through software that chooses between DoubleZero and the public internet. Austin Federa’s thesis: “We’re building this all on a broken foundation.”
- The near-term payoff is better Solana market structure, not faster-feeling retail clicks. A New York transaction may take 300 milliseconds or two seconds to reach a block builder in Singapore, exposing market makers to stale quotes and costly execution risk. Federa thinks predictable low-latency routing can drive slippage below 10 basis points, bring onchain pricing toward parity with Binance’s order book and return price discovery onchain.
- Validator economics are the adoption gate, and the early evidence is positive but modest. Roughly 20% of Solana stake was already connected at the time of recording; a six-month testnet reduced skipped blocks and lifted median rewards, though Federa stressed these were not “20% increases.” DoubleZero’s defining metric is whether it makes validators more profitable: “Today we are making folks slightly more money, and we hope that is more than slightly soon.”
- The token is compulsory network fuel with a burn-and-distribution loop, not an ornamental governance asset. Validators pay a seat fee targeted around 5% of block rewards—illustratively about 250 SOL per epoch worth of DoubleZero tokens, depending on the exchange rate—with roughly half burned and half paid to fiber contributors. The protocol currently has no inflation; future staking, if introduced, would carry what Federa described as a “very low” reward rate.
- The launch deliberately omitted a large retail airdrop and was scheduled to begin charging for live utility immediately. Federa argued that a backend protocol resembles an oracle more than a retail application such as Pump.fun, making a broad airdrop less natural. The investor thesis is therefore direct: transaction demand and execution speed must keep rising, validators must keep finding the network economical, and usage must become large enough for burns to reduce supply.
- An SEC no-action letter materially narrowed DoubleZero’s regulatory uncertainty, according to Federa. The request addressed token rewards for fiber contributors and eventual staking; his reading was that the SEC viewed the described activity as “fundamentally not regulated as a security.” He called it the first such letter for a crypto project in a very long time and a potential unlock for other compliant token networks, while acknowledging that discussing token value live still involved “legal landmines.”
- DoubleZero is base-layer neutral, but Solana supplies its clearest scale ambition: one million transactions per second within a few years versus roughly 10,000 today. Federa noted that Firedancer has benchmarked hundreds of thousands—almost one million—over private networks, suggesting the public internet is becoming the constraint. Thread Guy’s closing framing was: “The internet’s the problem now. We’re going to remove the internet from the problem.”
1. Solana’s FTX survival gave Federa permission to leave
Federa entered crypto in late 2017 at Republic Crypto, later worked at Bison Trails before its acquisition into Coinbase Cloud, and then spent four years as head of strategy at the Solana Foundation. He joined roughly nine months after mainnet and two months after Solana had laid off about half its employees—“not a consensus move back then.”
The FTX collapse arrived immediately after Breakpoint in Lisbon. Federa stayed in his hotel room for three days; for roughly eight subsequent weeks, he said the team slept only three or four hours nightly. The post-FTX response included an all-hands war-room call. Yet nobody internally thought Solana was finished because “the tech was super strong” and builders lacked anywhere comparable to go.
The validating moment came one year later, when another 4,000 people showed up for four days of Breakpoint programming. Federa left in November 2024 because Solana had survived FTX and the bear market: “Don’t have a big enough ego to think you got to stick around and help these guys.”
2. DoubleZero treats the public internet as crypto’s next bottleneck
DoubleZero is neither an L1 nor an L2. Its purpose is to create “a new internet optimized for high-performance systems like blockchain,” borrowing the dedicated-network model long used by Google, Amazon, Microsoft, Facebook and high-frequency trading firms.
The non-crypto precedent is brutally commercial. Facebook runs a giant private network so Instagram Reels load instantly and generate more ad impressions; Federa said it could wire up most of the Philippines with fiber to expand the market it could reach. Google built interconnected data centers to target a 50-millisecond response anywhere in the world. “The internet is too slow for how big these businesses have gotten.”
Twelve independent contributors provide DoubleZero’s fiber, spanning Jump, DRW, Cumberland, Galaxy, TeraSwitch, Latitude, Servers.com and participating venture firms. Unlike secretive trading networks, Federa described it as a consortium in which multiple contributors can operate on a shared system and experience higher quality.
Why now? Earlier blockchains were not fast enough to exploit dedicated infrastructure. With Firedancer, Anza, Aptos, Sui and high-performance L2s running into public-internet limits, DoubleZero’s premise is that the bottleneck is moving beneath consensus software into the physical network.
3. Faster routing is a market-structure product, not a convenience feature
The host’s pushback—worth keeping—was that Solana already feels instant to retail traders. Federa’s answer: its roughly 400-millisecond block time is “unfathomably slow” beside traditional finance, even though it is fast for crypto.
A New York trader first reaches an RPC, then whichever validator is building the next block—perhaps in Singapore. Public-internet conditions can make that journey take 300 milliseconds or two seconds. In the gap, Binance may move and a market maker can suddenly be selling an asset worth $1 for $0.90, losing hundreds of thousands of dollars.
High-frequency trading combines signal quality with reaction speed: a model might detect an iPhone-factory fire or tariff rumor, but the knowledge is useless if a competitor trades first. Federa’s signature framing: “Being faster in trading is basically as close as you can get to having a time machine.”
He rejected the idea that HFT is inherently market manipulation. Its useful function is pushing a displaced market back toward the “true price,” improving execution for retail users. For Solana, he sees a route to less than 10 basis points of slippage and onchain pricing comparable to Binance.
4. Network effects matter only if validators earn more
DoubleZero’s first users are validators, which can process transactions faster and eventually build larger, better blocks; the second are market makers and professional traders moving prices between Coinbase, Bybit or Binance and onchain venues such as Jupiter.
Benefits remain probabilistic while coverage is incomplete. If the next leader is outside DoubleZero, a cancellation still travels through the public internet; if connected, the trader reaches it faster. Federa even expects sophisticated firms might quote different spreads depending on which validator is scheduled to build the block.
Roughly 20% of Solana stake was already connected at the time of recording, before fees were scheduled to start the next day. A six-month testnet reduced skip rates—instances where a validator had the right to build but missed the block—and produced higher median rewards. Federa’s hedge was explicit: improvements are currently slight, “not like 20% increases.”
The core metric is therefore not raw bandwidth but validator profitability. DoubleZero is working with Jito and others to turn lower latency and additional capacity into more valuable blocks, including around a proposed Solana change—numbered 293 or 292 in Federa’s recollection—that would remove block compute limits.
5. The token converts validator demand into fiber economics
Validators pay in DoubleZero tokens for a network seat, with fees set around 5% of block rewards. Federa estimated approximately 250 SOL per epoch worth of tokens under full participation, while stressing that rewards, token exchange rates and resulting burn quantities are variable.
About 50% of fees are burned and 50% flow to fiber contributors, whose physical network costs run into tens of millions of dollars annually. The Cayman nonprofit foundation sits outside that transaction loop, much as the Solana Foundation receives none of Solana’s transaction fees.
The host pressed the uncomfortable implication: the foundation cannot retain validator revenue and must fund itself by selling tokens. Federa agreed, explaining that foundations eventually exhaust their holdings if token prices stop rising—but called that deliberate decentralization: “You don’t want a giant for-profit entity” owning the network’s economics.
There was no large launch airdrop. Federa contrasted DoubleZero’s backend utility with Pump.fun’s retail distribution: an airdrop can be effective for a consumer-heavy product, but “it’s kind of like an oracle airdrop” when the core users are validators and infrastructure operators.
6. The endgame extends beyond Solana—and down to the physical layer
DoubleZero is designed as base-layer-neutral infrastructure, with L2 tests and conversations underway across other L1s. Ethereum mainnet and Bitcoin are not built for performance, so Federa said they probably would not benefit greatly, though “flash boys” probably would; Hyperliquid could use the network for its read-node and state-propagation bottlenecks.
Multicast is one differentiated tool: rather than a Solana node transmitting roughly 2,000 copies of the same data, network switches replicate one packet stream and deliver it to every destination. Federa compared it to AirPlay sending one phone stream to ten speakers: “You can’t run multicast on the public internet.”
Physically, the network is interconnected fiber—“long strands of glass” carrying laser light through decades-old cables that remain useful today. Unlike public routes chosen partly to minimize cost and capable of bouncing New York-to-Tokyo traffic through several US cities, DoubleZero targets fixed paths and “functionally zero jitter.”
Validators connect through a software tunnel with access to both DoubleZero and the public internet, automatically selecting the better route for each destination. The engineering team—led by co-founder and CTO Andrew McConnell and staffed largely by networking specialists without crypto backgrounds—combines exchange, HFT, Amazon-networking and space-communications experience.
Federa’s closing target was one million Solana transactions per second within a few years, versus about 10,000 now and roughly 4,000 in 2023. Firedancer has benchmarked hundreds of thousands, almost one million, over private networks. Thread Guy summarized the wager as removing the internet constraint and giving Solana’s software teams a new set of engineering problems to solve.