What this covers
Kyle traces the scaling history of blockchains and uses it to place Solana against Bitcoin and Ethereum. He begins with Satoshi Nakamoto's one-megabyte block limit, which capped Bitcoin at roughly seven to ten transactions per second. The block size wars ruled out hard forks, leaving SegWit and Lightning, and Bitcoin gained only about 3x in six years. He then turns to hardware. Clock speeds stalled in the mid-2000s because of heat, so gains now come from parallelism, and the speed of light limits global consensus blocks to about 150 to 200 milliseconds.
The technical middle section explains how Solana responds. Each transaction declares which state it will touch, so non-overlapping ones can run in parallel execution on GPUs with thousands of cores, while the EVM uses one laptop core serially. Turbine spreads data through a known node hierarchy, and Kyle also describes why fees rise nonlinearly as utilization nears 90 percent. He argues that a finance-native chain needs low latency because leverage makes liquidations time-sensitive.
The contested material is on the Ethereum side. Kyle says automated market makers exist because fees of about 30 basis points and slow blocks ruled out order books. He claims optimistic roll-ups cut fees from about $50 to $2 but cannot reach 10 cents, and calls Sharding premature. Zero-knowledge roll-ups are his known unknown. He also faults the Ethereum Foundation for being hands-off and underpaying core developers, and argues credible neutrality is what justifies decentralization in DeFi. The episode closes on money: staking yield as a risk-free rate, and productive assets of roughly $800 to $900 trillion against about $10 trillion in gold.
Kyle argues Solana's parallel, finance-native architecture makes it the best-suited base layer for DeFi and non-sovereign monetary assets, while Ethereum's serial execution and optimistic roll-ups face hard scaling and fee limits.
- Serial execution in the EVM wastes most of a modern machine's cores, while Solana declares state access up front so non-conflicting transactions run in parallel
- Parallel silicon scaling gives a forecastable throughput path, whereas Bitcoin's 1 MB cap and Ethereum's serial design have stalled
- Optimistic roll-ups cannot reach sub-cent fees, so zero-knowledge roll-ups on a fast L1 are the likely route to mass scale
Parallel execution gives Solana high throughput and low fees that serial Ethereum and Bitcoin designs lack.
- Blockchain fees are nonlinear in utilization: they barely move up to about 50 percent load, rise modestly to 70 percent, and then increase roughly tenfold per marginal percent of demand near 90 percent, so growing demand worsens the experience for all users.
“when you get to 90% utilization, every marginal percent increase in demand fees go up 10 X.”
- Satoshi's arbitrary one-megabyte block size limit, added to stop spam, set a hard ceiling of roughly seven to ten transactions per second that Satoshi probably did not expect to persist.
“Blocks cannot be bigger than one megabyte. Super arbitrary determination.”
- Ethereum's Virtual Machine executes all transactions serially and its tooling assumes no parallelism, so a modern laptop uses one core and a GPU's thousands of cores go unused.
“So the EVM just runs everything serially. So your laptop today probably has four cores, maybe eight cores in it, and your graphics card is probably a thousand cores”
- Among next-generation blockchains, Solana is the only one attempting intra-shard parallelism, while other teams have instead pursued sharding.
“The only one that's really attempting to do intra-shard parallelism is Solana.”
- Solana enables parallel execution because each transaction header specifies which part of global state it will touch, so the system can run non-overlapping transactions simultaneously on GPUs and serialize only those with dependencies.
“each transaction header needs to specify what part of the global state it is going to touch”
- Solana transaction fees are around a thousandth or a hundredth of a penny, and Kyle argues a dollar-level fee would be unacceptable for large-scale business use.
“today Solana is in the fraction of a penny, like a thousandth of a penny or a hundred of a penny or something.”
- Writing parallel transaction processing is hard because low-level resource management is difficult, and few engineers know it, so Ethereum's team, lacking distributed-systems backgrounds, defaulted to serial execution as Windows and Mac took three to four years to support multi-core.
“it took three or four years for Microsoft and Apple to get Windows upgraded and Mac upgraded to support that.”
- Solana today runs at roughly 50,000 transactions per second, a large multiple of Ethereum's approximately 30 transactions per second and Bitcoin's roughly 7 to 10.
“today's Solana runs at, I'd call that 50,000 transactions per second.”
- Solana's ability to run parallel execution comes from an execution-layer design choice rooted in its team's high-performance and distributed systems background, not from its consensus algorithm or proof of history.
“this is just a function of designing an execution layer for writing code that has nothing to do with consensus algorithms or proof of history”
- Solana runs transactions natively on modern NVIDIA graphics cards with thousands of cores, so a card with about 4,000 cores can execute thousands of jobs in parallel, versus the handful of cores a laptop uses for serial code.
“Solana runs transactions natively on graphics cards, modern NVIDIA cards that I'd call have 4,000 cores”
- The only known solution to safe parallel transaction execution is for each transaction to declare in advance which parts of global state it will touch, allowing non-overlapping transactions to run concurrently and overlapping ones to run serially.
“each transaction header needs to specify what part of the global state it is going to touch”
- Vitalik Buterin's 2014 Ethereum introduction at the Miami keynote showed no clear idea what the platform was useful for, and he could not coherently articulate DeFi despite alluding to concepts in the video.
“he has no idea what this thing is useful for”
Solana can scale to meet large future demand as a fast, finance-native layer one.
- Reed Hastings started Netflix streaming before bandwidth was available by projecting bandwidth's growth curve, and Kyle argues blockchain builders can similarly bet on silicon trajectories.
“Reed Hastings wanted to do streaming at the time of starting, but knew that the bandwidth wasn't there”
- Because of the speed of light traveling around the planet, global consensus blocks cannot realistically be produced faster than about 150 to 200 milliseconds, which forces systems to decouple time from communication.
“When given the inherent latency of the speed of light going around the planet, you really can't get blocks under 150 milliseconds”
- Sharding gives one parallel thread per shard, but cross-shard communication adds latency and gas costs, so the key question is how far a single shard can scale before adding shards.
“kind of the key questions I think about scaling these systems is can you scale a shard.”
- Blockchains break software's zero-marginal-cost assumption because the world's storage, compute, and bandwidth are fixed in the short term and shared among all users, so metering via fees is required.
“in a blockchain, you have a fixed amount of resources, you have three basic resources. You have storage, you have compute, and you have bandwidth.”
- Within a window of seconds to a minute, most transactions in a block do not depend on one another, so they can be executed in parallel, with chronology mattering only for the minority that touch shared state.
“The vast majority of transactions that probably happen within a block... probably don't have dependencies on one another.”
- CPU clock speeds stopped rising around 2005 to 2008 because heat dissipation grows super-linearly with clock speed, so computing gains now come from parallelism and specialization, making throughput growth forecastable.
“The reason it hasn't increased is because this heat creation and therefore the need for heat dissipation, increases super linearly with clock speed.”
- Kyle is concerned that demand for Solana could outstrip its capacity by a large factor within five years, even though a single shard has room to grow for several years.
“I actually am pretty concerned that demand will outstrip supply for Solana by a large factor in the next five years.”
- Solana's design reflects its founder Anatoly's two decades of high-performance hardware and operating-systems work at Qualcomm, Dropbox, and others, which led him to design a distributed network to run as fast as possible.
“His whole life, he spent saying, How do I take an existing piece hardware and make it go as fast as possible?”
- The highest-probability path to billions of daily users producing around a hundred billion transactions per day is maxing out Solana layer one and then extending it with zero-knowledge roll-ups.
“I think the only way we're going to get there is zero-knowledge stuff on top of Solana.”
SOL's utility, burn, and predictable monetary policy make it a durable long-term holding.
- SOL has value from four sources: it is required to pay gas fees, its burn mechanism could make it scarcer over time, it is portable trust-minimized collateral, and its native staking rate serves as a risk-free rate for the network.
“So one is, you have to use these to pay gas fees... Three is a very trust minimized source of collateral... And then four, you have a native staking rate”
- Solana's inflation rate starts near 7 percent and decreases over ten years to roughly 1.5 percent, and Solana has burned a portion of each transaction fee since launch, similar to Ethereum's EIP-1559.
“The inflation, I believe was set like 7%, and then it decreases over the first 10 years down to something like one and a half percents.”
- Global wealth is dominated by productive assets such as debt, equity, and real estate, totaling roughly 800 to 900 trillion dollars, compared with about 10 trillion in gold, a ratio of roughly 100 to 1.
“gold is like 10 trillion or something, then all of the other wealth in the world... debt equity and real estate... 800 trillion, 900 trillion”
- Non-sovereign stores of wealth are fundamentally defined by predictability of monetary policy, which Kyle thinks needs to be within about 1 percent annualized, rather than by fixed supply or simplicity.
“what matters as you think about non-starters where wealth is really just predictability of monetary policy. And you need to get that to probably within 1% annualized”
- Solana is Multicoin's largest position, bought at prices from about 4 cents up to 140 dollars, and the fund has sold almost none of it and intends to continue holding.
“Solana's probably our largest position. We own a lot of it. We've sold almost none of it from 4 cents to $140”
- Proof-of-stake networks provide a native staking rate that functions as the risk-free rate of that system's currency, analogous to government bond yields for fiat, though the speaker acknowledges risk is not zero.
“you have a native staking rate, which is effectively access the risk proof rate in these systems”
- A base-layer token is valuable as a commodity needed to pay gas, as a potentially deflationary asset because of fee burning, and as trust-minimized collateral that can be levered anywhere, and the speaker argues it is actually more than a triple point asset.
“I actually think it's probably more than a triple point asset”
- What matters for a non-sovereign store of wealth is predictability of monetary policy, and it needs to be held to roughly 1 percent annualized deviation, and the speaker expects Solana could match Ethereum's neutrality within two to three years.
“you need to get that to probably within 1% annualized, it's probably fine.”
- For investors in crypto, liquidity can be a bug rather than a feature because it makes it too easy to sell during drawdowns, and the speaker's firm held SOL through painful internal debates about selling.
“I definitely appreciate common thinking these days, which is liquidity is a bug, not a feature.”
Consensus and propagation mechanics let Solana coordinate nodes more efficiently than alternatives.
- Proof of history is not a consensus algorithm but a sequential hashing process that gives each node a verifiable passage of time, allowing the network to order transactions and rotate leaders without all nodes communicating in real time.
“proof of history is not a consensus algorithm”
- Proof of work is technically a Sybil-resistance mechanism, while the consensus rule Bitcoin and Ethereum use is Nakamoto Consensus, where the chain with the most accumulated proof of work is authoritative.
“That consensus algorithm that Bitcoin and Ethereum currently use... is called Nakamoto Consensus.”
- In Solana, every node counts hashes locally, so after about a million hashes (roughly 400 milliseconds) leadership rotates, and a non-responding leader is skipped automatically without any shared clock being maintained.
“everyone is running their million and hashes, they all just know to skip you, and it just goes to the next person.”
- Solana's Turbine protocol propagates transactions along a known hierarchy of nodes, like BitTorrent, so data spreads in logarithmic rounds rather than random rebroadcast, which is the fastest known approach for broadcasting to many nodes.
“Solana uses a networking protocol as they call Turbine... it looks a lot like a BitTorrent actually.”
- Bitcoin and Ethereum network gossip protocols are resilient to random node failures because nodes rebroadcast to peers, but they are inefficient because redundant random rebroadcasting wastes bandwidth.
“The great thing about gossip protocols is they're extremely resilient... The downside of gossip protocols is they're extremely inefficient.”
- Proof of stake chains use classical Byzantine fault tolerance consensus families, the most popular being BFT, which requires two-thirds of stake to agree on state before finalization.
“the most popular of those is called Byzantine fault tolerance or BFT”
Crypto's most exciting frontier is real-world applications and an active Solana ecosystem.
- The most exciting category for crypto is re-imagining industries using crypto-native assumptions and forms of capital formation and coordination, which Kyle believes will produce hundreds of new companies and protocols.
“area number three is the area I'm most excited about.”
- Solana Labs has about 70 to 80 employees and is likely to reach around 200 within a year, but is not expected to exceed 1,000.
“I think Solana Labs is 70 or 80 employees today. I'm guessing a year from now, it'll be 200 people.”
- Helium is a new business model in which individuals deploy wireless hotspots and are paid in tokens for localized coverage, replacing centralized carriers and tower owners, and it moved to Solana after failing on Ethereum.
“Helium is a new business model for deploying and managing wireless networks.”
- Bitcoin's early token incentives rewarded unproductive work such as running patches and mining coins, whereas Helium is the first version where the rewarded activity creates real-world value.
“the do stuff was not productive. It was run patches and mine some coins. It's kind of useless for the world.”
- Creators are under-monetizing, so the temptation to experiment with token-based monetization is too high to avoid, and the question is whether it happens in the US or elsewhere.
“I think they all know that they're under monetizing. And so I do think that given that reality, it has to happen.”
- Solana Labs built Metaplex, an NFT infrastructure platform supporting storage, resale fees, and related features, and gave it away free, reportedly 2,000 times in 30 days, to lower developer friction.
“Solana team built it and gave it away 2,000 times in the last 30 days.”
Finance requires low latency and on-chain order books to reach sub-basis-point spreads.
- Automated market makers on Ethereum worked only because high fees of about 30 basis points and slow block times made on-chain order books infeasible, so the AMM model was the viable substitute.
“Uniswap worked because Ethereum was slow. Specifically, it worked because Ethereum was slow.”
- The central limit order book is the foundational pillar of finance because the price of an asset is set by bids and offers that market makers post, and spread-crossing trades produce price discovery.
“What is an asset worth? The answer is: what people are willing to sell it for, what people are willing to buy it for.”
- A finance-native blockchain needs very low latency and high throughput because derivatives introduce leverage, leverage creates blowout risk, and managing liquidations requires fast execution.
“anything in finance that has derivatives means you have leverage. If you have leverage, you have risk of blowouts. And if you have risk of blowouts, you need to have low latency.”
- Blockchains should be reframed not as programmable non-sovereign money but as the best conceivable DeFi system that also has non-sovereign monetary properties, which implies the design should be written from the ground up to be finance-native.
“reframe blockchains as the best conceivable DeFi system that happens to have non-sovereign monetary properties to it”
- The future of finance is trading at sub-one-basis-point spreads rather than 30 basis points, which an on-chain order book on a fast chain can enable.
“The future of finance is not 30 basis points trading; it's sub one basis point trading.”
Credible neutrality, not censorship resistance, is why decentralized DeFi and on-chain markets matter.
- The killer app for blockchains is DeFi, interpreted broadly to include recreating all financial contracts with auditability, composability, and instant settlement, and extending financial concepts into non-financial domains.
“the killer app for blockchains is DeFi. And I think you should probably interpret DeFi as broadly as possible.”
- The best proxy for credible neutrality is the number of parties involved in a system and how evenly they share influence over it, since a million parties with equal power cannot act against anyone.
“Probably the best proxy for credible neutrality is the number of parties involved in the system and degree to which they have influence over it.”
- Audius moved to Solana because logging likes, plays, and social graph events for about 5 million monthly actives in a credibly neutral way is impossible on Ethereum given its throughput.
“At 5 million monthly actives, there's no chance this would ever work on Ethereum.”
- The core reason decentralization matters in DeFi is credible neutrality, meaning every party can trust the transaction layer will not change in ways that favor someone, which is a more important property than censorship resistance.
“I think it's probably less about censorship resistance and more about credible neutrality”
- Serum is a generalized on-chain framework for running a central limit order book that can power spot trading, derivatives, interest rates, and other markets, built by the FTX and Alameda team after Sam Bankman-Fried found Ethereum could not host an order book.
“Serum was built by the team at FTX in Alameda, which are run by Sam Bankman-Fried”
On-chain ownership of tokens and NFTs unlocks creator monetization and new real-world benefits.
- Social tokens can be issued with a programmatic bonding curve in which each purchase raises the price along a formula and deposits reserve funds that keep the coin solvent for buybacks, giving creators built-in liquidity.
“the price of a coin starts at X, and then every time someone wants to buy the coin, when you buy the coin, the dollars go in the reserve”
- Blockchains let creators tie a share of revenue, such as 10 percent of artwork sales, to automatically buy their own coin, which can be proven programmatically on-chain to supporters.
“10% of the revenues for all my artwork sales are going to flow into buying my coin, right? And you can literally prove that programmatically on-chain.”
- On-chain ownership of in-game assets in Star Atlas enables third parties to extend the game state without developer permission, a capability Kyle thinks requires a Solana-class chain at Eve Online scale.
“all the assets in the game are owned by someone... any third-party could build extensions onto the game”
- AMC's offer of free popcorn to stock owners is an early example of using ownership to unlock real-world benefits, an idea Kyle expects companies to adopt more widely within five to ten years, though the verification process was poorly executed.
“that's the first example of a much bigger idea in capital markets and capital formation of using ownership to unlock real world things”
- NFTs combined with social tokens will be widely adopted within five years, first by the TikTok, YouTube, and Twitch creator class and then spreading outward.
“I think they're going to be very, very widely adopted in the next five years, primarily by the TikTok, YouTube, Twitch class”
Solana's underdog strategy rejects Ethereum conventions and gives away foundational primitives.
- Solana is currently less credibly neutral than Ethereum but is becoming more neutral as its ecosystem and number of parties grow.
“Solana is at the current moment, I would say less credible neutral than Ethereum, but it is becoming more credible neutral.”
- Solana succeeded by leaning against Ethereum-era conventional wisdom, rejecting the claim that TPS talk is unserious and skipping EVM backward compatibility, which made it a focal point for people who disagreed with the masses.
“by leaning strongly against the status quo, what you end up doing is you become a shelling point”
- Solana Labs builds foundational primitives such as Serum and Metaplex and gives them away without monetizing them, a strategic difference from Ethereum that Kyle thinks is necessary for an underdog.
“Solana team built it and gave it away 2,000 times in the last 30 days.”
- The Ethereum Foundation has been hands-off in building ecosystem infrastructure and underpays core developers, a strategy Kyle argues has been value destructive relative to what Ethereum could have been.
“they've basically built no infrastructure around the Ethereum, they have basically underpay or don't pay core developers at all.”
Sharding should wait until a single chain is maxed out, since layer-1 design is well understood.
- The probability of unknown unknowns that would overturn layer-1 blockchain design is below 5 percent because distributed systems, cryptography, and programming models are well understood.
“I think the probability as it pertains to layer-1 blockchains that there are unknown unknowns at this stage is pretty low, like sub 5%.”
- Sharding does not make sense until a single shard has been pushed to its capacity, so pursuing it now is out of order and creates solvable but annoying problems.
“Sharding by definition, doesn't make sense until you have maxed out the paths where you have the shard.”
- Sharding only makes sense after a single shard is maxed out, so Solana should first reach around 10,000 transactions per second on one shard, and only then consider a path to a million transactions per second.
“Sharding by definition, doesn't make sense until you have maxed out the paths where you have the shard.”
Optimistic and zero-knowledge rollups lack a credible path to low-cost, general, large-scale use.
- Optimistic roll-ups reduced Ethereum fees from about 50 dollars to about 2 dollars but have no credible path to 10 cents, so they are not intellectually interesting for large-scale use.
“Optimistic roll-ups I find to not be intellectually interesting at all. They get the fees down from call it $50 to call it $2 on Ethereum.”
- Zero-knowledge proofs are hard to build and reason about because few engineers understand them, no academic programs teach them, and they are not fully programmable or composable, creating a real chance they never become general.
“very, very few people know how it works... There's a chance that it's never good enough to be fully general”
- Zero-knowledge rollups are the biggest known unknown for scaling because few engineers understand them, they are not fully composable or programmable, and it is uncertain whether they will become general-purpose within 12 to 84 months or ever, though they would run better on Solana than Ethereum.
“The biggest known unknown is zero-knowledge roll-ups for sure.”
Bitcoin's scaling has underdelivered, while its protocol remains more neutral than Ethereum's.
- The block size wars resolved against hard forks, so Bitcoin's only scaling routes became data compression, as with SegWit in 2017, and off-chain Lightning, which has grown poorly; Bitcoin gained only about 3x in six years.
“the side that said you can't introduce the hard fork that breaks the rules of the system. And that camp kind of won”
- Bitcoin maximalists' insistence that a store of value must be simple and unchanging is a backward-looking view that mirrors gold, and analogizing physical-world properties to digital ones is almost always wrong.
“it's just fundamentally a backwards facing view of the world”
- Bitcoin is more credibly neutral than Ethereum because its unchanged protocol for six years leaves no party able to meaningfully alter it, whereas the Ethereum Foundation controls the roadmap and makes political decisions.
“Bitcoin is definitely more neutral than Ethereum because the reality is that Bitcoin hasn't changed in six years.”
Holding a volatile token with conviction is worthwhile despite temptations and regulation.
- Solana's genesis supply was 500 million tokens, with the largest allocation a community bucket of roughly 20 percent controlled by the Solana Foundation, which Kyle thinks will likely be burned rather than spent.
“Solana's genesis block was 500 million tokens.”
- Holding a volatile token through conviction is harder than it looks because liquidity tempts sales, but this is a feature of long-horizon investing rather than a flaw in the asset.
“I definitely appreciate common thinking these days, which is liquidity is a bug, not a feature.”
- Even if the probability that Solana becomes the most important financial infrastructure in the world is only 10 to 20 percent, holding it at current prices is still wildly positive expected value.
“call it even 10% or 20%, you don't have to assign a 50% or 60% probability to that, but if you do, then it is still worth holding at the current price?”
- Social tokens may face one of three regulatory outcomes: they move offshore, they grow so fast that the SEC must reconcile them, or the SEC gives a safe harbor, which Kyle considers unlikely under a Democratic administration.
“One outcome is this doesn't happen in the United States... another case that this becomes so widespread and so popular so quickly... A third is just, that it kind of happens slowly”