Hardware as Ideology: An Interview with Austin Federa

Introduction

Blockchains don’t run on hope. They run on cables, consensus, and the quiet conviction of people who build systems the world can’t see. 

Austin Federa has spent most of his career moving between the visible and the invisible. In marketing and communications—from his early days at the Boston Globe to shaping narratives on Solana—the architecture is rhetorical, laying beams of meaning, wiring them with story, and hoping the structure holds against the ever-watching eyes of the public. In hardware, the architecture is literal: silicon, topology, and latency, where the real work happens beyond view. 

In an era where digital empires rise and fall on the strength of unseen foundations, hardware isn’t just machinery—it’s ideology incarnate. It transcends mere circuitry to silently dictate how we connect, compute, and create. Federa was a key architect of how the fastest blockchain on the market explained itself to the world, navigating turbulent times as Head of Strategy at the Solana Foundation. Now, as co-founder of DoubleZero, he’s turned from messaging to mechanisms, building a new purpose-built Internet for the next era of distributed systems.

I wanted to understand what happens when a storyteller becomes an architect—when the person who shaped a protocol’s public identity turns to shaping its private backbone. This interview is about that shift—from narrative clarity to network clarity.

The conversation has been edited and condensed for brevity.

Interview

Origins and Worldview

Ichigo: When you were fifteen, you caused quite a stir at Apple by getting your Dell laptop to run a protected build of OS X. Was that your first real foray into tinkering with hardware? What was it about hardware that grabbed your interest?

Federa: Yeah, maybe this is kind of a good starting place for me: I’ve got a medium to mild form of dyslexia, and that made things like learning a foreign language really, really difficult. I tried to learn Spanish in high school, and it just… didn’t click. In college, I ended up taking Latin because, honestly, it was easier—you weren’t expected to speak it, the work was translation. That kind of thing worked better for me.

The same thing happened with programming. There are parts of it I can get, and parts I can understand, but I’m just not cut out to be a great coder. If you give me a blank coding window, I just… it’s just not going to happen. But hardware felt different. Especially at fifteen, when your brain is still kind of developing. It felt more grounded, more structured. There are fewer variables, quite frankly, compared to software.

So yeah, back in middle school, I was building my own gaming computers. I even had a little—I wouldn’t even call it a business—but I built and sold a few systems to other folks. They were water-cooled rigs, back when that was actually hard to build. Today, you can just order the parts online, but back then, you really had to figure it out.

I was always pretty good at managing software that was applied to a specific problem or assembling other people’s code into something useful. But hardware was the thing that felt natural, where I could really put the pieces together. That’s what grabbed me.

Despite this early interest, you took an academic detour. You studied political science, environmental studies, and economics—fields not typically associated with hardware. What drew you to those areas, and how did they help shape how you think about things today?

When I was younger, probably in early elementary school, I was a really science-brain kid. Like, very into science. I went to this unusual school—not exactly Waldorf, but with academics plus a big property for exploration. Recess was fun, damming up streams, causing all sorts of problems. It was a different experience, and my head was really into that science space back then.

As I got into middle school and high school, I started exploring the humanities more. That became interesting to me, something I really enjoyed, and by the time I got to college, that’s what I wanted to do. I went to a liberal arts school, and I didn’t go in with any aspirations of being an engineer.

Ironically, a lot of my friends did—most of them went into engineering degrees, even at liberal arts schools. One of the guys who works with me now, Ben, actually went to high school with me. He ended up at Dartmouth for engineering. So even in that circle, I was surrounded by people going deep into technical fields, while I was off studying political science, environmental studies, and economics—that’s just a blockchain degree.

For me, engineering never felt “easy,” but the problems seemed more straightforward. The humanities and social sciences felt messier, more open-ended. That’s what I was drawn to.

Quite frankly, I think if you can have the luxury of doing it, college is the only time in your life that you are without judgment and have the courage to just explore your curiosities in an academically rigorous way. So, I took advantage of that.

In your bio, you’ve got this line, “Everything’s Computer.” This mantra suggests a worldview that hardware isn’t just a bunch of tools, but something more foundational to reality. How did early experiences like the Apple hack and then your academic background shape this philosophy?

I guess I didn’t fully understand a lot of this until maybe three or four years ago. I was talking to Kevin Bowers—and not to bring this back full circle to Solana too quickly—but one of the things I always felt was that the definition of hardware and software is kind of arbitrary.

We’ve moved into this phase where you don’t even think about the hardware—you just run stuff on virtualized instances. On one level, that’s elegant, but on another, it feels… not how computers actually work. Going back to the Hackintosh project, for example, that whole thing was about instruction sets. I had to trick a Dell CPU into thinking it had instructions that it didn’t, and sometimes it would just hard crash. That was one of those moments where you realize the line between software and hardware isn’t as clear as we think.

I also grew up reading a lot of sci-fi—books like Accelerando by Charles Stross—where a single breakthrough in computation changes the trajectory of everything. That shaped how I think about systems: we need to pay more attention to what’s going on at the low level, because that’s where the real leverage is.

A lot of this was just years of ad-hoc learning I’ve been doing over the years related to hardware. My general philosophy is that if you learn enough around a topic, you’ll eventually figure it out. Structured, step-by-step textbook learning never really worked for me—I’d always feel like I was missing something. But if you keep circling around the problem, eventually the deeper picture comes together. 

It’s kind of funny, but math class never really clicked for me the way it did for some people. It was difficult in part because it assumes we’re teaching a very basic version that we’re going to go deeper and deeper. It wasn’t until I was much older that I was like, “Oh, everything is totally wrong and that has nothing to do with how math actually works.” It’s just the way we’ve been teaching it for a few hundred years. We miss out on some foundational stuff, and math becomes infrastructure. It’s a reminder that even the abstract stuff eventually grounds out in how systems actually run.

Oh, I agree one hundred percent. When I sat down and taught myself the math behind ZK, I wondered why I was never taught half of this material. Even just trying to read it, you have all these weird letters, and then you realize that weird “E” is just a for loop. There are a whole host of knowledge assumptions that create communication issues.

From Communications to Protocols

This is a nice segue because, for a long time, you were focused on communications. Communication professionals obsess over clarity, whereas hardware engineers obsess over latency. Do you see any parallels between building performant hardware and building performant narratives?

Kind of. I’d say the biggest difference is that both systems have to work, but the technology side has to work in a much more literal sense than the communication side does.

On the narrative side, you can build a story about a product or a candidate or a piece of code that doesn’t actually reflect what it does. Ripple is a classic example—people would say, “Well, the banks will all need it someday.” And it's like, okay, maybe… but what’s the actual path to that outcome?

You see this all the time in society, too. The COVID slogan flatten the curve is a great example. Super sticky as a narrative, but flattening the curve isn’t an end state—it doesn’t solve COVID. Same thing with protest slogans like ceasefire, now. A ceasefire is, by definition, temporary. It works as a rallying cry, but it doesn’t capture the whole picture. Narratives can be incredibly sticky even when they don’t line up with the underlying reality.

The same thing happened with ultrasound money. It never really made sense technically, but it gave people something to rally about. So narrative structure tends to be very different from technical product structure.

That said, you do need to think about the story you want your product to tell before you build it. Amazon has that famous line that you should write the press release before you write the product brief. It’s been around forever, but people forget it. You have to start with intentionality. I think a lot of engineering does not begin with intentionality. It begins with a series of discrete problems that someone’s attempting to solve, and you end up with a bunch of solutions pieced together instead of something more coherent.

Honestly, that’s why the best engineering often happens during wartime. In those moments, you start with hard requirements: we need a plane that can get from the runway to 40,000 feet in less than ten minutes, have this range, and fly at this speed. It doesn’t matter if the plane looks ugly—it meets the requirements, and you build a thousand of them. That’s a very different kind of engineering than working on problems without a clear story of what you’re trying to achieve.

It sounds like much of the work in creating performant narratives or hardware is almost invisible by design. How do you make something visible enough to attract developers, but invisible enough to feel seamless for the end user?

I’d argue very few things that actually change the world fall into that invisible-by-design category. When the Internet came on the scene, for example, it made a huge splash. People had to learn about phone lines and weird acoustic couplers—hard products to use, but powered by transformational tech. 

You can say the same about GPUs. When NVIDIA launched CUDA, it wasn’t exactly a household name, but among engineers it became one. Everyone knew Sun Microsystems for similar reasons. Those moments weren’t just optimizations—they were step-function changes.

Compare that to switching hardware vendors: Cisco versus Juniper versus Arista. Important differences, sure, but mostly incremental. What gets public attention is when something is so disruptive you can’t ignore it. That’s why everyone knows ChatGPT and, earlier, Bitcoin.

So to me, it depends: if you’re building an optimization, maybe it doesn’t need to be “seen.” But if it’s a true breakthrough, the visibility will happen on its own. People learn about it because it does something that fundamentally wasn’t possible before—like SpaceX catching a rocket with a pair of chopsticks.

Speaking more towards those disruptive changes, you worked for Solana Labs and later the Solana Foundation. And before that, you had more of a marketing and communications background, from the Boston Globe to Bison Trails. How did you find your way to Solana? What brought you in?

Yeah, so that’s a good question. I graduated from college in 2013, and I thought about whether I wanted to do a PhD program or not. For me, I was in this phase where I was looking for something a little bit more practical.

I remember going on the NBER website, where all the economics papers get posted. Someone had actually done a paper about the NBER itself, which is like the most meta social sciences thing you could do. I looked at it, and there were three authors, and it was only downloaded five times. You’re just kind of like, wow, like nobody reads these things at all. Like, is this work really making a difference?

Journalism was kind of like a stopover for me. I believe civic engagement is very important for society, and that requires a certain amount of knowledge and awareness. Working at the Boston Globe gave me a chance to see how narratives get created in real-time, and I was able to help get out a lot of information to help people come up with an opinion of the world and be more informed. It was the first place I thought I could really do some work here. 

But, I found that a lot of other people I was working with were not particularly good in the sense that they didn’t take their job with a huge mission behind it. I don’t want to be working in a place where people don’t seem passionately engaged with their work.

That’s part of what pulled me toward startups. In the startup world, people are obsessed—sometimes unreasonably so—but you don’t question whether they care. Everyone’s moving fast, everyone’s trying to get something off the ground. That energy was really different from what I’d felt before.

So I moved from journalism into communications, then into campaigns, consulting, and eventually tech. My first real full-time job in crypto was working for Republic and launching Republic Crypto, which is very much on the investment side. Then I moved over to Bison Trails, a blockchain infrastructure company that ended up becoming Coinbase Cloud. There are some really awesome people who came out of that organization.

And so, you know, I did not want to work for Coinbase—it just didn’t seem like the right thing. Bison Trails was getting acquired, I wasn’t going to have a job soon, and I was interviewing with a bunch of Layer 1s and Layer 2s.

To take a step back, Bison Trails was running about 65% of the Ethereum 2.0 network back then during the Beacon Chain phase. I was deep into the ETH 2.0 roadmap back when it was going to launch with 56 shards, and they were going to scale to a thousand shards before they scrapped that part of the roadmap.

I kept having this thought: how is Uniswap going to work in a sharded ecosystem? Are we going to have a Uniswap shard? An Aave shard? How is this all going to work? There was a lot of handwaving around bridges, and we’ll figure it out. You know, all these types of things. I kept asking a version of this question to all the founders I was interviewing with. Everyone was doing some form of sharding. 

I remember asking Toly how shards are going to work out. He laughed and said they’d never work—you need everything in one global state machine. This was the first credible take I’d seen about what the future of shards will actually look like, and it wasn’t a bunch of handwaving software people saying we’d figure it out later. That was the moment when I was like, I have no idea if Solana is going to be successful or not, but I’m going to learn a ton here.

Building DoubleZero

Of course, your time at Solana speaks for itself. You eventually left the Solana Foundation in December 2024 to co-found DoubleZero. Was there a single “aha” moment that convinced you to start a new protocol focused on network infrastructure?

There wasn’t any one moment. We got through the FTX collapse, and that was an all-hands-on-deck situation for a while. But by early 2024, the Foundation was doing good stuff, but it was becoming less relevant to the ecosystem, which is a sign of success, as the Foundation’s influence declines.

The work there was shifting into more of a maintenance period, continuing the mission. People like Dan Albert describe it as needing folks around for stability over the long term. For me, the question was: do I want to do that, or try my hand at founding and building something new?

I didn’t want to build something competitive with Solana or just another application on it. Nothing wrong with that, but going from running strategy at Foundation to building something like a DeFi protocol would create a dual-master problem—it’d be hard to let go of the old role.

When thinking about becoming a founder, you have to ask: Why am I doing this? Is it for the aesthetic, or something else? 

A lot boils down to: are you better at this than anyone else currently trying? Startup founders aren’t the best in the world—if they were, they’d be elsewhere. It’s about that mix of opportunity, calls, and timing.

The Firedancer conversations convinced me that Solana needed to go beyond the public Internet to reach its potential. That’s where we set out to build DoubleZero. 

For those unfamiliar, can you break down DoubleZero’s two-ring architecture? What are the overarching problems it solves? I’m interested in how DoubleZero decided on this architecture. What alternatives were considered?

You can think about it as: what are blockchains really bad at? I like to start with systems, not in what they should be doing, but what they’re doing that isn’t good. Blockchains are bad at propagating state and sharing resources.

On sharing resources: if you’re running an app on AWS or Google Cloud, you have their entire cybersecurity infrastructure protecting you—the data center is your defense layer. The goal with DoubleZero is to leverage resource sharing, making any overwhelming attack a problem for the data center or ISP.

Solana can get degraded by DDoS attacks that wouldn’t touch Netflix or Bank of America. If we deploy defense layers in front of every validator, it’s a one-to-one relationship. But one FPGA defense box can protect dozens of downstream validators. Without that, you’d need $20,000 FPGAs in every validator, which is untenable—and it doesn’t help connectivity.

The two-ring architecture comes from two desires: 

1. To protect validators from spam, invalid signatures, and garbage
2. Speed up the networking layer 

Speed up networking without defense, and you’ve just made a better DDoS tool. Add defense without speed, and you’re only making it more resilient, not faster. We want a million TPS on mainnet with 2,000 nodes worldwide—that’s where the fiber comes in. 

Aside from sharks, what are the biggest technical risks or open questions DoubleZero needs to face to help get to that fated one million TPS with 2,000 nodes worldwide?

There are a number of them. This is something I wish were more acceptable in the industry. A lot of people don’t talk about the problems they still need to solve. For us, once you move into physical infrastructure, it requires humans plugging in cables and other sorts of little things that could trigger protocol-level events.

One of the main things for us is testing links for integrity. In Web2, you’d hire a third-party to plug devices in, measure, and give a cable certification on speed. We, for obvious reasons, don’t want that, so we’re building a system where software on my fiber runs a speed test on yours. It’s a verifiability problem that happens in the real world. Because in software, you have the beautiful qualities of signatures and cryptographic security, but none of that applies to the physical world at all.

One of the challenges we face in this space is how do we do this in a way that is as trustless as possible, but also delivers a hundred times the performance that you would get over the public Internet. One of the nice advantages we have is that there are no funds stored on the DoubleZero network. It’s not a smart contract network. In that sense, we don’t have custody over user funds, and so we don’t need the type of economic security models you do, or people think you do, on Ethereum or Solana. 

There’s a whole bunch of other stuff we have to do. We’re also building out prioritized data, like HFT-style Flash Boys for blockchain, which will be cool.

Future Vision and Personal Reflections

Looking ahead, what’s on the roadmap for DoubleZero in the next one to two years? What would you like to accomplish?

The number one piece is getting as much stake as we can on the network, so protocol developers can push limits further. The Holy Grail: within two years, Solana running at least 100,000—or a million—TPS. If not, that’s a loss.

For sure. A lot of this conversation has been centered around blockchains. What’s one non-blockchain use case you’re excited to see, or want to see, be implemented on DoubleZero?

We already have a bit of this, which is pretty cool. Shelby is a storage system competing with S3 and Cloudflare R2, built on DoubleZero. It’s a token project, but at its core, it’s decentralized fast flash storage. It’s interesting because now we have tons of applications that can actually compete with centralized systems.

There’s also interest in AI training and workload data—that’s harder but exciting. OpenAI is actually one of the biggest subsea fiber users linking data centers, so there’s definitely something there. 

Looking back, you’ve worn many hats: tinkerer, communicator, strategist, founder. What’s one thing you wish you had known earlier in your career?

When I was younger, I focused too much on what people thought I should do or society’s expectations. There are good components to that—a lot of what’s wrong with modern influencer culture is people lacking moral or philosophical grounding. But you have to be willing to take risks.

Nothing is as long as it seems. Coming out of college, a U.S. senator advised me to go to law school. At the time, I wasn’t cut out for it, but I should have reconsidered later. I don’t regret not going, but it was a path someone said I could do, and I didn’t believe them. In retrospect, that could have been cool.

Rapid Fire Questions

What’s the most “invisible” technology that you admire?

Building control systems—the industrial-grade HVAC stuff that keeps buildings at perfect temperature and humidity. It doesn’t get enough love, but it’s a prime example of tech improving lives invisibly.

What’s the hardest human or organizational problem in building infrastructure that lasts?

It depends on the level. Usually, incentive design: infrastructure failures stem from differing time horizons or budget assumptions. In the physical world, like roads, it’s from not understanding the system upfront—the foundations weren’t built for the volume. In software, complacency is the threat. 

Look at United Launch Alliance versus SpaceX: ULA did everything by the book, but based on false safety assumptions driven by NASA’s risk aversion. Elon flipped the script by embracing failure in a private structure. 

The structure that’s building something is not usually appreciated enough for how it influences the outcome. It’s like that old saying: “Show me an org chart and I’ll show you your organizational priorities.” 

What’s the most overhyped thing in Solana right now?

I’m not sure anything is overhyped at the moment. Six months ago, I would’ve said memecoins, but now the hype feels proportional. 

I do think people are being far too short-minded around the United States, though. I am a big believer that the U.S. can become the crypto hub of the world. But, like, I see protocols that are giving up their international foundation to become a 504 or a DUNA (Decentralized Unincorporated Nonprofit Association). And, man, imagine we get President AOC in 2028? That’s going to look pretty bad in retrospect. 

People in the U.S. and in Solana are very bullish on the current administration. I have high hopes for them and think they’ve done a really awesome job, especially with the changes in the SEC. But for the changes that aren’t passed through Congress, they can just be undone in the next administration, right? We have to create durable change.

What’s your favorite piece of hardware you’ve ever owned?

Probably my old Fuji XT2 camera—a mirrorless digital with analog, tactile interfaces. It was a joy to shoot on, and nothing else was built that way back then.

What’s your go-to civ in Sid Meier’s Civilization 7?

Yeah, haha, this is not going to be a rapid-fire answer—you’ve opened a can of worms that could last for hours.

Hmmm, I think seven was really rough when it came out, and it got a lot of hate, but it’s turned into a really solid, well-structured game. At this point, there are some simplifications in some of the mechanism designs that I’m not happy with. I miss era points and the influence that you can have, taking over people’s cities through culture. 

In general, I tend to go for a more science-oriented leader. In Civ 7, I’ve actually been having a lot of fun with the Carthage-Caesar civilization and leader sort of team-up, which is just like a full power military domination play, and then you really struggle to recover from that in the second age because you overexpand pretty quickly.

When DoubleZero mainnet?

Our goal is before mid-October. It’s all engineering-dependent.

Conclusion

As we trace Austin Federa’s path—from a teen hacking hardware to a strategist narrating Solana’s ascent, and now pioneering DoubleZero’s subterranean networks—we uncover a profound truth: the invisible often begets the revolutionary. 

What began as rhetorical scaffolding evolved into silicon sinews, where latency isn’t just a metric but a frontier to be conquered in the digital era. 

Federa’s reflection reminds us that true innovation demands intentionality, whether that be in creating sticky narratives or forging unbreakable links. Complacency corrodes, and hype fades without durable roots. 

The future is optimistic: DoubleZero eyes mainnet by mid-October, hoping to power a million transactions per second in the coming years. Hardware emerges, not as an auxiliary, but as an ideological core, propelling us towards seamless, scalable, sovereign systems.

This conversation isn’t merely about Federa’s foray into founding DoubleZero—it’s a blueprint for builders everywhere: embrace the unseen, question the arbitrary, and architect realities that endure. 

The quietest cables carry the loudest convictions.