Why Three Chains Exist

This Part opened with a promise in the overview: to answer the question the reader brought to the book — why do BTC, ETH, and SOL all exist if they’re “all just blockchains”? Now we can answer it properly. They exist because they are solving different problems, and each design pays a different cost to do it. They look like competitors, but they’re better understood as three points on a single map — and the shape of that map is the blockchain trilemma.

The trilemma: you can’t max all three

Popularized by Ethereum’s Vitalik Buterin, the blockchain trilemma observes that a single-layer chain struggles to fully achieve all three of these at once:

• Decentralization — many independent participants can verify and produce blocks on modest hardware.
• Security — attacking the ledger is ruinously expensive.
• Scalability — high throughput, low latency, cheap transactions.

                    DECENTRALIZATION
                         /\
                        /  \
                       / BTC\          push toward any one corner
                      /      \         and you trade away another
                     /  ETH   \
                    /  (+ L2s) \
                   /            \
        SECURITY  ◄──────────────►  SCALABILITY
                          SOL

You don’t get to ignore the trilemma — you get to choose where to stand on it. Each of the three chains made a different choice, and everything else about them follows from it.

BTC — optimize for trustless soundness (store of value)

Bitcoin plants itself in the decentralization-plus-security corner and refuses to trade either away for throughput. The base layer stays small enough that almost anyone can verify the entire chain; the rules change rarely and conservatively; the supply is fixed and computable by every node. The product that falls out is money — a maximally credible, censorship-resistant store of value and final-settlement layer, graded honestly in monetary properties.

• Problem it solves: a sound, apolitical money whose scarcity and rules no one can quietly change.
• Cost it pays: very limited base-layer throughput and expressiveness — everyday payments and anything programmable get pushed to layers like Lightning and the scaling frontier.

ETH — optimize for programmability (world computer)

Ethereum spends some of Bitcoin’s monetary minimalism to buy expressiveness: a global, shared computer where anyone can deploy programs (smart contracts) that strangers can call and trust to run as written (see smart contracts & the EVM). To keep that computer decentralized without choking the base layer, it exports scaling to rollups rather than growing the L1.

• Problem it solves: trust-minimized applications — money that can also enforce arbitrary agreements.
• Cost it pays: more complexity and a larger attack surface, plus a base layer kept deliberately modest, so cheap throughput depends on the L2 ecosystem maturing.

SOL — optimize for throughput (one fast state machine)

Solana takes the scalability corner seriously and bets that one monolithic, high-performance state machine can serve everyone directly — no rollups, near-zero fees, sub-second slots (see Solana architecture and fees, throughput & blockspace).

• Problem it solves: high-frequency, low-cost applications that want one fast, unified global state.
• Cost it pays: heavier validator hardware (a decentralization cost) and, historically, lower resilience — the network has halted under load where Bitcoin has not.

The map, on one page

Chain	Optimizes for	The product	What it trades away
BTC	Decentralization + security	Store of value / settlement	Base-layer throughput & expressiveness
ETH	Programmability (kept decentralized)	World computer / app platform	Simplicity; cheap throughput leans on L2s
SOL	Scalability / throughput	One fast global state machine	Validator decentralization; some resilience

By the numbers — the trilemma as verification cost

The trilemma is abstract until you price it in who can afford to verify. Roughly, as of 2024–2025:

   Bitcoin full node:     hundreds of GB of disk, runs on a hobbyist PC / Raspberry-Pi-class box
   Ethereum full node:    heavier (fast SSD, more RAM), still home-runnable by enthusiasts
   Solana validator:      high-end server — lots of RAM, fast NVMe, high bandwidth

The cheaper it is to run a verifying node, the more independent checkers exist, the harder the ledger is to capture — which is exactly why Bitcoin guards that number jealously and Solana accepts a higher one to go fast. Same trilemma, three prices.

Under the hood — why one chain can’t just “do it all”

The binding constraint is that every full node must independently re-verify the chain. Push throughput up and each node must process more per second, so weaker machines drop out and verification concentrates — spending decentralization. Hold decentralization fixed and you cap how much work the base layer can take on. L2s are the escape valve: do the heavy work off-chain, but post just enough to L1 that a cheap node can still check it. That’s why the honest disagreement isn’t “who’s fastest” but where to put the work so the verification stays cheap.

It actually happened — the cost of conservatism, and of speed

Bitcoin’s refusal to move fast is itself a design choice with a track record. In March 2013, an accidental incompatibility between two Bitcoin software versions briefly split the chain into two histories; miners coordinated within hours to converge on one, and Bitcoin has run without an unplanned network-wide halt since. That episode hardened Bitcoin’s “change almost nothing” culture. Solana sits at the other end: optimizing for throughput, it has suffered multiple multi-hour outages (notably ~17 hours in September 2021 and ~5 hours in February 2024). Neither history is a verdict — they’re the visible bills for opposite choices on the same trilemma.

They’re points on a design space, not strict rivals

It’s tempting to ask “which one wins?” — but that assumes one problem with one right answer. There isn’t one. A maximally sound, never-changing money and a fast, flexible application platform are different goods with different costs, and a design that’s ideal for one is, by the trilemma, compromised for the other. The three chains coexist for the same reason cash, a programmable bank, and a high-frequency exchange coexist: they’re optimized for different jobs. Borrowing between them is common — Bitcoin experiments with more expressiveness, Ethereum and Solana work hard on credibly decentralizing — but the centers of gravity stay distinct because the underlying tradeoff is real and permanent.

The thread — three answers to one question

The whole book has circled one question: how do untrusting strangers come to agree on one ledger? Now you can see that BTC, ETH, and SOL are three different answers to it, each minimizing a different cost.

• Bitcoin answers by making the ledger so cheap to verify and so resistant to change that strangers need trust nothing but arithmetic — and pays for it in throughput.
• Ethereum answers by letting strangers agree not just on balances but on programs, so a contract can be the neutral party — and pays for it in complexity.
• Solana answers by making one shared state machine fast enough that strangers can transact in real time — and pays for it in how many of them can independently keep the network honest.

That’s why three chains exist. Not because the market hasn’t “picked a winner,” but because agreeing on a ledger without trust is a problem with several reasonable solutions, and these three sit at different corners of the same map. Knowing the map — the trilemma and its costs — is what lets you reason about the next chain the same way, instead of asking which logo to cheer for.

Check your understanding

1. State the blockchain trilemma’s three properties and the core claim about achieving them on one layer.
2. For each of BTC, ETH, and SOL, name the corner it optimizes for and the main cost it accepts.
3. Why is “node verification cost” a good concrete proxy for where a chain sits on the trilemma?
4. Explain why a single base layer “can’t just do it all,” using the requirement that every full node re-verifies the chain.
5. In what sense are the three chains not strict competitors? Use the idea of a design space.

Show answers

1. Decentralization (many can verify on modest hardware), security (attacks are ruinously expensive), and scalability (high throughput, low cost). The claim: a single-layer chain can’t fully maximize all three at once — pushing toward one corner trades away another.
2. BTC optimizes decentralization + security (sound money), trading away base-layer throughput and expressiveness. ETH optimizes programmability while staying decentralized (world computer), trading away simplicity and leaning on L2s for cheap throughput. SOL optimizes scalability/throughput (one fast state machine), trading away validator decentralization and some resilience.
3. Because the cheaper it is to run a verifying full node, the more independent participants can check the chain, which is the essence of decentralization. Bitcoin keeps that cost low (hobbyist hardware); Solana accepts a high-end-server cost to gain speed — so verification cost directly reveals the tradeoff each made.
4. Every full node must independently re-verify the chain, so raising throughput forces each node to do more work per second. Weaker machines drop out and verification concentrates (losing decentralization); holding decentralization fixed caps base-layer throughput. Hence one layer can’t max both — L2s exist to move work off-chain while keeping L1 checking cheap.
5. They solve different problems with different costs — sound money, a programmable platform, and a fast unified state machine are distinct goods, and by the trilemma a design ideal for one is compromised for another. So they’re different points on one design space, like cash vs a programmable bank vs a high-frequency exchange — coexisting because they’re optimized for different jobs, not racing for a single prize.