Timechain
Search Donate
Sponsors
Cake Wallet, private Bitcoin and Monero walletSideShift.ai, no-signup crypto exchangeOrangeRock, non-custodial bitcoin spot and perps tradingBraiins Hashpower, buy and sell Bitcoin hashrateLayer Two Labs, drivechains for Bitcoin
S16 · E30 · 5:23:41

S16 E30: Peter Rizun on Scaling Bitcoin with Hardware

13 Jun 2025 · by Vlad Costea · 13 min read
Published at block 901,110
Cover art for S16 E30, Peter Rizun on Scaling Bitcoin with Hardware
MP3 (445 MB)

The audio keeps playing while you browse the site.

Read the transcript50,692 words, every line links to the minute it was said

Ask this episode

Embed this episode

Paste this on your site or blog. It is a small player that streams from here: no trackers, no cookies, no ads.

Preview the player

Peter Rizun and I went live for 5 hours and 23 minutes, with more than 1,600 people watching at the peak, which is a lot of strangers to keep company with while a Chief Scientist of Bitcoin Unlimited explains elliptic curve multiplication. By the time we got to his hardware, three hours in, I had to say it out loud: "we went on for three hours now and we haven't touched on the hardware scaling stuff that you're working on." The first three hours had gone to the block size war, and neither of us seemed bothered.

Peter has been in that war since 2013. Around then he was paying for beer in Vancouver with 0.1 BTC, and he told me some of his friends lost the keys to coins that would now be worth 14,000 CAD. I remembered those years as a time when I woke at 5:00 AM to watch Asian market trading patterns, and he remembered the same thing, so we spent the first stretch comparing early-Bitcoin scars before anyone said the word megabyte. The question underneath the whole show was his, and he stated it plainly: "we want to create a P2P e-cash that the entire earth can use, all the people in the world can use."

One Gigabyte and the Arithmetic of Earth

For Peter the target is a number. "We just have to scale till we get to that Earth scale that we're shooting for, and that's around one gigabyte blocks." He cites Dan Kaminsky's 2011 slides as the technical roadmap, and he does not think the debate in 2015 was ever about hitting that number on day one. "Well, two megabyte blocks would have been better than one," he said, and he prefers BIP101, which starts at 8 MB and doubles every two years. When I asked why not just do something simple, his reply was short: "I think the simpler thing is just make the block size bigger."

I wanted to see the math on the table, so we ran it. A year of 1 GB blocks is about 52 TB, which Peter prices at roughly "a thousand dollars per year every year to store the block chain." On bandwidth I tripped over my own units and he corrected me. Streaming a 1 GB block takes "13.3 megabits per second," and the relay bandwidth for a node that already has most of the transactions in its mempool is far lower, "only 4.8 megabits per second." At one point I said something about terabyte blocks, and he cut in: "You're talking terabyte blocks, we're talking gigabyte blocks before."

Storage and bandwidth are the easy half. The hard half, he said, is state: "I think that's the hardest part of scaling Bitcoin, is scaling the state." The UTXO set has to live in fast memory, and his estimate was "One terabyte of RAM would cost maybe $3,000 at today's DDR4 prices." That is the bottleneck he spends his time on, and it is why the hardware work comes later in this article.

I pushed on the node-count objection, the one every small-blocker raises: if a node costs a fortune, nobody runs one. Peter recalled the video Peter Todd commissioned, the one with "these huge data centers" where "nobody can run a node because they cost like a billion dollars," and said the funding for it came from the same circle that leaked the "John Dillon" emails we also touched on. Dillon, the pseudonym in those emails, funded his small-block video and also pushed the CoinJoin bounty on Dark Wallet, and the two of us sat there reading the coincidence back to each other. His own picture of a healthy network is modest. "Most people are using SPV wallets and maybe one out of every hundred hobbyists who's a super keener like me has their own node. That is a that's a win for Bitcoin." He also draws a line at the other extreme: "you don't want to like say a node cost a million dollars in order for the transaction to be one cent."

Who Forked, and Who Stopped the Talking

I asked about the thing every old big-blocker has to answer sooner or later, which is whether hard forks are as safe as he makes them sound. His position on the history is that the change that mattered came before any fork. "all hell broke loose because it was no longer permissionless," he said of the moment the block size became something you had to ask for. And then, in his telling, "they started censoring people like suddenly you're not allowed to talk about raising the block size limit." On his own side he is willing to take blame too: "we shot ourselves in the foot," he said, which referred to the strategy of the camp he belonged to. Bitcoin Unlimited had a bug in 2017 that disabled nodes, and he acknowledged it without my having to drag it out.

My pushback was on the word consensus. If a hard fork needs it, I asked, how do you know you have it? He agreed it can be fuzzy ("I guess you kind of know it when you see it, maybe") and he held the other half of the position firmly: "if you don't have consensus, you always have a risk of a fork." That was where we kept circling, because I think the fork risk is the whole cost and he thinks it is the price of an open protocol.

On SegWit he is less hostile than I expected. "I didn't like SegWit, but I was okay with it," he said. He also said "they were just trying to help us move forward," about the people who proposed it, which is not how that period is usually remembered by either camp. Where he thinks his side lost the plot was strategy: "We thought the most important people to go after were the miners, but really the most important people to connect were the exchanges." He also said "that one will become the real Bitcoin," which is how he described the chain that exchanges list under that name, and I pointed out that this is a political fact about exchanges rather than anything the code decides. In my words, "human politics, you know, whereas code is code." He answered with a concession and a conviction at once: "I guess I can't prove that I'm right, but that's what I believe."

Turing Completeness and a Physicist's Satoshi

Peter's complaint about Bitcoin Script is one he has held a long time, and he gave it to me directly: "if Satoshi had got the scripting system right to begin with, we wouldn't need these new opcodes." He would keep the UTXO architecture and add "a Turing-complete scripting language, maybe like RISC-V using a RISC-V virtual machine." The point of that, in his words: "Everyone can just build whatever they want because everything is possible in the native architecture of Bitcoin." He cites Ethereum as proof of the premise ("what we've seen is that Turing completeness is really useful. I think Ethereum has proved that. Smart contracts lead to all sorts of uses that we can't even predict, and it allows for permissionless innovation") while holding that Ethereum "shot itself in the foot by using an account-based model."

That led us, as it tends to, to OP_CAT. I made a joke about cats and he asked, deadpan, "So it doesn't involve real cats?" It does not, and I had to explain the opcode to the one man on the stream who thinks the whole category of opcodes is a bandage. I also said what I think about the era: Satoshi's emails with Hal Finney read to me like Hal "patronizing him, treating him like he's a lesser coder." Peter disagreed, and what he said was about Satoshi's mind, not Hal's manners: "To me, Satoshi has a mind of a physicist." Where I see a coder getting talked down to, he sees someone reasoning from first principles about a physical process, and the thing that would have occurred to him about proof-of-work was the wonder of it: "Look at all these leading zeros in these block headers. That must have taken like the enormous computational power to solve these hash puzzles. Why did that happen? That would be so interesting."

Why He Still Criticizes Lightning

The first thing he said when Lightning came up was almost a mission statement: "I still criticize Lightning to this day." His technical claim is narrow and specific: "I think Lightning is broken in a high fee volatility environment." When on-chain fees spike, channels cannot be closed or defended cheaply, and the safety of the whole design leans on being able to get to the chain. He had a second argument that is legal rather than technical, about routing nodes: "to me, they obviously are money transmitters because that's their entire purpose."

I pushed back with the usual defenses. Peter's answer to the scaling pitch was that it only works if everyone has channels with everyone, and he put the consequence bluntly: "every single connection between every node of the network is going to be a little payment channel." He is not volunteering to fix any of it: "I'm not going to work on that because I have a lot of other things to do." Part of that, I think, is that he sees Lightning as a rival answer to a question he already answered with bigger blocks, though I did not get him to say so in those terms.

The thing that does bother him in general terms is where the payment rails have ended up. "If our future is one of stable coins or central bank digital currencies, then I think that would be a terrible, terrible outcome," he told me, and later he went further: "most payments are now stable coins. And I think that's really sad because to me that's like a backdoor way into getting central bank digital currencies." Paul Sztorc, who has been on the show a number of times, has been attacking Bitcoin's direction from a different angle for years, and the idea that Bitcoin should be forked to get what it needs is one he and Peter would both recognize.

The Fee Market, Blitz, and Nexa

The economics of fees are where Peter claims to have a mathematical edge. "we've now proven that there is not this infinite demand for block space," he said, which comes from the orphan risk research: a miner who builds a bigger block pays a cost in orphan probability, so supply of block space is priced even without a cap. I wanted to know what he made of today's filter debates, and his answer was layered. "I see the transaction fee as the primary filter, but I think it makes sense for nodes, including mining nodes, to set parameters on what is garbage." He also pointed to F2Pool, who return accidental transaction fee overpayments voluntarily, as evidence that "a lot of the reason society works is because we have etiquette," and said "most people aren't scammers."

Then, finally, the hardware. Blitz is an FPGA on an M.2 card, and Peter describes it plainly: "this will do around 200,000 signature verifications per second at around three watts of power." That works out to "one microjoule to verify a transaction." The card has no CPU in the loop: "there's no processor running code." The reason he wants it is that he does not trust himself, or anyone, to run a full node on a general-purpose machine: "I don't want to run Bitcoin or whatever, or Nexa, whatever it is on my main desktop." The thing he wants to cheapen is the node. "it'll turn nodes into a commodity," he said, and his joke about who loses from that came a few minutes later: "So I want to fire Core."

I asked about plans. His current goal "is just to ship this eventually, which will be some signature checks as part one." There is a catch he named himself: "a bit of a chicken and egg problem," meaning no one builds the software until the hardware exists and no one buys the hardware until the software exists. His way around it is Nexa, where proof of work is tied to the work nodes do: "We're trying to tie proof of work to the bottlenecks in running a node." It uses elliptic curve multiplication in the hash puzzle, so a card that accelerates the curve math also mines. I asked him straight why he is building it, and he gave me the unvarnished motive: "Well, the answer why I'm developing it is because it's going to mine Nexa better." On progress he said "Right now Nexa is, I would say it's maybe 25% of the way there." If other people adopt the hardware for other purposes, "that's a win for me too."

Eric Voskuil was on the show for S5 E8 on Libbitcoin development, cryptoeconomics and Austrian economics. Peter's version is built around hardware instead of Eric's software architecture.

Bitcoin Cash, Craig Wright and One Owed Bitcoin

I pushed him on the viability of Bitcoin Cash. Calin Culianu, Steve Thurmond and the Phenomenal Big Blockers later recorded S17 E26 conversation with Calin Culianu, Steve Thurmond and the Phenomenal Big Blockers on nine years of Bitcoin Cash. He did not move: for him the measure is whether the chain scaled, and Bitcoin Cash did not.

"I think Craig Wright really damaged the big block movement and Bitcoin Cash," he said, and "He owes me a Bitcoin." The origin is a bet on selfish mining math. Kurt Wuckert Jr later recorded S17 E39 on the big block wars, Craig Wright and the road not taken. Peter and I did not get into it.

He also used Amaury Séchet's eCash fork as an example of what he dislikes about arbitrary governance, where the block reward is a lever someone can pull. And when I asked what the excitement of the early days looks like now, he answered flatly: "But yeah, all of that excitement's gone." Asked about the community's attachment to Roger Ver-style personalities, he said "And people love this guy." He did not say who he meant at that moment, and I left it there.

On the future of his own career he drew a line I was not expecting: "There's no point in my work there," he said of anything that is not about cash, and then "That's not why I got involved. I got involved for peer-to-peer electronic cash." The way he put the goal at the start of the show was the same thing in a longer form: "the goal of Bitcoin when Satoshi created it, and my goal today is to help build better money for the world. And for that, I think we needed to scale to earth levels."

Wasps, a Maybe-Who, and Never Saying Price

Late in the show we drifted to Ross Ulbricht's speech in Las Vegas, the one about decentralized wasp nests, which Peter and I both found moving and neither of us could explain very well. I asked who might be behind some of Bitcoin's open mysteries and he said "Definitely not Mr. You-Know-Who, but I don't know," which got a laugh out of chat. Then, with the stream near its end, I tried to get a price prediction for Nexa. His answer was five words and has now been committed to the record: "Never say price until it's listed."

I mentioned that Vancouver, where Peter lives, has largely stopped accepting Bitcoin at shops, which is a sad change from the beer-for-bitcoin days. He just shook his head. After five hours we were both tired, the chat was still going, and the last thing either of us wanted to do was explain to a stranger why a coin was worth something.

Listen to Peter Rizun on Spotify, Apple Podcasts, YouTube, and More

Time stamps:

  1. 00:01:15– Introducing Peter Rizun
  2. 00:03:32– Early Block Size Debates and Satoshi’s Vision
  3. 00:07:45– Block Size Limit History and Miner Soft Limits
  4. 00:10:10– Dan Kaminsky’s 2011 Scaling Predictions
  5. 00:12:01– Scaling Philosophy: Zero, One, Infinity and Earth-Scale
  6. 00:15:09– Block Size Wars, Chain Splits, and Altcoin Proliferation
  7. 00:16:43– Stablecoins, Lightning, and Bitcoin’s Use Case
  8. 00:19:03– Zcash, Ethereum, and Bitcoin’s Missed Opportunities
  9. 00:22:11– Bitcoin’s Script Limitations and Smart Contract Innovation
  10. 00:25:10– Scaling Limits: Hardware, Storage, and Bandwidth
  11. 00:35:31– Node Software, Formal Specs, and Core’s Dominance
  12. 00:44:00– Censorship, Reddit, and the Block Size Debate
  13. 00:48:04– SPV Nodes, Custodial Wallets, and Decentralization
  14. 00:51:30– Block Size, Internet Speed, and Global Node Access
  15. 00:57:26– UTXO Model, Dust, and State Management
  16. 01:01:37– Ethereum, Zcash, and Node Benchmarking
  17. 01:03:26– Hardware Acceleration and Specialized Chips
  18. 01:07:00– Sponsorship Break and Bitcoin Adoption via Debit Cards
  19. 01:14:59– History of Block Size Proposals and Client Forks
  20. 01:21:07– Consensus, Forks, and the Role of Exchanges
  21. 01:25:40– SegWit, SegWit2x, and Political Compromises
  22. 01:33:04– Bitcoin Cash, Market Dynamics, and Altcoin Competition
  23. 01:39:34– Stablecoin Fees, Global Demand, and El Salvador
  24. 01:46:52– Decline of Bitcoin Payments and Missed Opportunities
  25. 02:10:56– Lightning Network: Promise vs. Reality
  26. 02:22:12– Shitcoin vs. Bitcoin Maximalism and Open Source Innovation
  27. 03:05:52– Spam, Miner Policy, and Transaction Filtering
  28. 03:17:05– Hardware Scaling: M.2 Accelerator and Node Commoditization
  29. 03:29:11– Nexa: Experimenting with Proof-of-Work and Node Bottlenecks
  30. 03:37:44– Craig Wright, Satoshi Identity, and Big Block Movement Damage
  31. 04:02:03– Fork Proliferation, Community Division, and Stablecoins
  32. 04:04:31– Bitcoin’s Future, Experimentation, and Open Research

Verify this episode

This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.

SHA-256
16ae5e7ee1d6289119c9db2437b8b0d34bc266fc29607584e82d074326330fff
Proof
2025-06-14_s16_e30_peter_rizun_on_scaling_bitcoin_with_hardware_-_14.06.2025_16.39.mp3.ots · all episode hashes (proof)
Status
Anchored in Bitcoin block 974,312
Timestamped
After release. The episode came out on 14 Jun 2025, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
  1. Download the MP3 and the proof into the same folder.
  2. Run sha256sum 2025-06-14_s16_e30_peter_rizun_on_scaling_bitcoin_with_hardware_-_14.06.2025_16.39.mp3 (on a Mac: shasum -a 256 2025-06-14_s16_e30_peter_rizun_on_scaling_bitcoin_with_hardware_-_14.06.2025_16.39.mp3). The result must match the SHA-256 above.
  3. Install the OpenTimestamps client (pip install opentimestamps-client) and run ots verify 2025-06-14_s16_e30_peter_rizun_on_scaling_bitcoin_with_hardware_-_14.06.2025_16.39.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node, ots info 2025-06-14_s16_e30_peter_rizun_on_scaling_bitcoin_with_hardware_-_14.06.2025_16.39.mp3.ots shows the block height to look up in any block explorer.

Share: X · Email · Link

Related

Bitcoin Cash Big Fork Blocks BCH Research Magazine

Bitcoin Cash: The Big Block Breakaway

19 Sept 2026 · 22 min read

Bitcoin Cash is the primary consequence of Bitcoin’s 2015-2017 block size war. It produced Bitcoin’s first durable chain split…

Comments

Comments run on my own server. No trackers, no tracking cookies.

0:000:00